covalently functionalized carbon nanotubes and their biological
Post on 30-Dec-2016
230 Views
Preview:
TRANSCRIPT
SCUOLA INTERNAZIONALE SUPERIORE DI STUDI AVANZATI INTERNATIONAL SCHOOL FOR ADVANCED STUDIES
Covalently functionalized carbon
nanotubes and their biological
applications
Thesis submitted for the degree of Doctor Philosophiae
Candidate Francesca Maria Toma
Advisors Prof. Giacinto Scoles Prof. Maurizio Prato
Coadvisors Dr. Loredana Casalis
Dr. Tatiana Da Ros Dr. Marcella Bonchio
October 2009
Table of contents
i
Table of contents
Acknoledgments .............................................................................................................................................. 1
Abbreviations .................................................................................................................................................... 3
Thesis abstract ................................................................................................................................................. 5
Chapter 1: General Introduction and aim of the thesis ...................................................... 9
1.1 Nanotechnology: Small Can Be Big! ........................................................................................... 9
1.2 Carbon Nanotubes ............................................................................................................................ 11
1.2.1 Synthesis and Functionalization of CNTs ................................................................... 16
1.2.2 Applications and Properties of CNTs ............................................................................ 19
1.3 CNT toxicity ........................................................................................................................................... 22
1.4 Aim of the thesis .................................................................................................................................. 23
References ....................................................................................................................................................... 24
Chapter 2: Covalent functionalization of carbon nanotubes ...................................... 27
2.1 Synthesis and Physicochemical characterization of a Carbon
Nanotube−Dendron Series ................................................................................................................... 27
2.2 Microwave-assisted functionalization of carbon canotubes in ionic liquids 40
References ....................................................................................................................................................... 51
Chapter 3: MWCNT dendron series for efficient siRNA delivery ............................. 55
3.1 Drug delivery in cancer nanotechnology: siRNA ............................................................ 55
3.2 CNTs and Dendrimers as drug carriers ................................................................................ 58
3.3 MWCNTs dendron series for an efficient siRNA delivery .......................................... 59
3.3.1 (G2)dendron-MWCNT:siRNA complexation ............................................................. 61
3.3.2 Effect of the different generation alkylated dendron-MWCNT:siRNA
cellular uptake ........................................................................................................................................ 63
3.3.3 Alkylated (G2) dendron-MWCNT:siRNA delivery and silencing with a
cytotoxic sequence ................................................................................................................................ 66
Table of contents
ii
3.2.4 Alkylated (G2) dendron-MWCNT:siRNA delivery and silencing with the
use of a cytotoxic sequence ............................................................................................................ 68
3.2.5 In vivo biodistributional study of dendron-MWCNT derivative ................... 69
References ....................................................................................................................................................... 73
Chapter 4: CNT Substrates for neuronal cultures: a morphological study ....... 77
4.1 Introduction: a new standard for substrate preparation .......................................... 77
4.2 Morphological study of neurons deposited on CNTs ................................................... 85
4.3 Pairs of neurons electrophysiology ........................................................................................ 90
References ....................................................................................................................................................... 94
Conclusion .......................................................................................................................................................... 95
Appendix A: Review of the experimental methods used in this thesis ................ 99
A.1 Carbon Nanostructures Functionalization using 1,3-dipolar cycloaddition
reaction ............................................................................................................................................................ 99
A.2 Microwave .......................................................................................................................................... 102
A.3 Thermogravimetric Analysis and Kaiser Test ............................................................... 104
A.4 Raman Spectroscopy .................................................................................................................... 107
A.5 Visible-Near Infrared Spectroscopy .................................................................................... 110
A.6 Trasmission and Scanning Electron Microscopy ......................................................... 111
A.7 Atomic Force Microscopy .......................................................................................................... 114
A.7.1 Electrostatic Force Microscopy ..................................................................................... 116
References .................................................................................................................................................... 117
Appendix B: Experimental section .............................................................................................. 123
B.1 Materials .............................................................................................................................................. 123
B.2 Syntheses ............................................................................................................................................ 124
B.3 Characterization techniques and sample preparation ............................................. 131
References .................................................................................................................................................... 137
Acknowledgements
1
Acknowledgements
My thanks are due first to two people that represent the best guides for studying
science that I could ever have found: Prof. Giacinto Scoles and Prof. Maurizio Prato.
They have directed me towards many different collaborations and given me the
opportunity to work in an international environment, encouraging me to take an
interdisciplinary approach.
I would like to thank Prof. Scoles for his guidance and encouragement, for always
making time to offer me advice, to teach and discuss things with me, and above all,
for trusting me from the beginning and for giving me innumerable opportunities to
learn new things.
Many thanks to Prof. Prato for his invaluable help, knowledge and understanding
and for allowing me to experience such a stimulating atmosphere. I would also like
to thank him for his gentle and constant presence and for finding the time to see
me whenever I asked him.
Special thanks go also to Dr. Tatiana Da Ros, Loredana Casalis and Marcella
Bonchio for their enthusiasm, help, useful meetings and invaluable discussions. I
have benefited greatly from their precious advice.
Many thanks to all the people I have worked and collaborated with, especially Dr
Andrea Sartorel and Mauro Carraro from Padua University, Dr Khuloud Al-Jamal
and Prof. Kostas Kostarelos from School of Pharmacy in London, Mr Matteo Iurlo
and Prof. Francesco Paolucci from Bologna University. Thanks to Dr. Alberto
Bianco for his availability and helpful discussions, Dr. Marco Lazzarino for enabling
me to use TASC laborotories and to Glaudio Gamboz for teaching me to use TEM.
Acknowledgements
2
Thanks also to Prof. Laura Ballerini for electrophysiological studies and especially
to Miss Giada Cellot for sharing her neurons with me!
I owe many thanks to all the past and present members of Prof Prato’s group and
SENILab for the enjoyable experience of working and having good time with them!
Thanks to Marian for her cooperation and help. Thanks to Coke for his valuable
remarks.
I would like to thank all the people who I have met over this three years, who have
enriched me in so many different ways. Thanks to GiokLan for her unlimited
kindness and hospitality.
Thank to the SISSA and SBP sector for giving me this precious opportunity.
Special thanks to Carla and Barbara for their company and for making nice my
experience in Trieste! Thanks to Denis, Fouzia, Nadja, Joceline, Anu, Tomas, Alessia,
Sara S., Fulvio, Pietro, Loredana, Mauro C., Mildred, Chiara, Sara C., Valeria, Niksa,
Alan, William and Tatiana for sharing with them unforgettable time.
As always, I want to thank Sara and Marco for their lasting friendiship and support,
for their constant presence despite the distance.
My really heartfelt thanks to my parents for their support and encouragment
through my years of studying, for their extreme love, and to my sister for her
useful tips. Thanks to Mauro for his extraordinary patience, for his help and
pointers and for lovingly taking care of me!
Many thanks to all of you! I am extremely grateful!
Francesca
Abbreviations
3
Abbreviations
AFM: atomic force microscopy
[bmim]BF4: 1-butyl-3-methylimidaziolinium tetrafluoroborate
CNTs: carbon nanotubes
CT: computed tomography
CVD: chemical vapor deposition
DMF: N,N-dimethylformamide
DMSO: dimethylsulfoxide
DOS: density of states
dsRNA: double strand RNA
DTPA: diethylentriaminopentaacetic acid
f-CNTs: functionalized carbon nanotubes
EFM: electrostatic force microscopy
E-EFM: enhanced-electrostatic force microscopy
EPR: enhanced permeability and retention
FET: field effect transistor
FT-CNTs: fluorous-tagged carbon nanotubes
G band: graphite band
G0: zero-generation
G1: first-generation
G2: second-generation
GEF: growth enhancing factors
HiPCO: high pressure CO
[hvim][(CF3SO2)2N]: 1-hexadecyl-3-vinylimidazolinium
bis(trifluoromethylsulfonyl)imide
Abbreviations
4
ILs: ionic liquids
MW: microwave
MWCNTs: multi walled carbon nanotubes
o-DCB: ortho-dichlorobenzene
[omim]BF4: 1-methyl-3-octylimidazolinium tetrafluoroborate
PAMAM: polyamidoamine
PSCs: post synaptic currents
RBM: radial breathing mode
RES: reticuloendothelial system
RISC: RNA-induced silencing complex
RNAi: RNA interference
SEM: scanning electron microscopy
siRNA: small interference RNA
SPECT: single photon emission computed tomography
STM: scanning tunneling microscopy
SWCNTs: single walled carbon nanotubes
TEM: transmission electron microscopy
TGA: thermogravimetric analysis
TUNEL: terminal deoxynucleotidyltransferase–mediated nick end labeling
UV-Vis-NIR: Ultraviolet-Visible-Near Infrared
Thesis abstract
5
THESIS ABSTRACT
In this thesis covalent functionalization of carbon nanotubes (CNTs) is
presented as a powerful tool to allow the use of CNTs for biological applications.
Covalently functionalized CNTs (f-CNTs) have a promising future as drug
carrier1. In our group, we are currently investigating this opportunity, especially
towards multiple functionalization of CNTs to load at once different compounds2
(drugs, solubilizing agents, imaging payloads). To address this issue, we prepared
new CNT derivatives using a dendron moiety with a high number of free amino
groups, which could be eventually used to prepare compounds with at least two
different payloads.
We have also studied alternative techniques of functionalization, exploiting
both thermal heating and microwave (MW) irradiation3, and succeeding in
functionalizing them with a solubilizing agent (i.e. tetra-alkyl ammonium salts) to
increase the aqueous solubility of f-CNTs4.
Taking into account the unique ability of carbon nanotubes to penetrate cell
membranes, the positive charges on the derivatized multi-walled CNT (MWCNT)
surface make them potentially ideal vectors for small interference (siRNA)
delivery. siRNA needs to be translocated into cells because, bearing many negative
charges, it cannot cross cell membrane by diffusion. Although viral vectors are
widely used, they imply many safety risks and involve inflammatory and
immunogenic effects. This is why non-viral vectors are preferable and many
cationic derivatives, such as cationic polymers, dendrimers and peptides, are the
most quoted candidates to translocate siRNA (and other genetic materials) across
cell membrane5. Using a fluorescently labeled, non-coding siRNA sequence, we
demonstrated that cytoplasmic delivery of the ribonucleic acid is remarkably
improved with increasing dendron generation. Furthermore, we demonstrated
Thesis abstract
6
also the possibility to translocate toxic-siRNA sequence exploiting this system.
These results demonstrate that the new dendron-based derivatives are very
promising substrates for application in nanomedicine.
Moreover, we have also investigated the possibility to use CNTs as substrates for
neuronal cell growth. We first functionalized CNTs to make them soluble in order
to produce CNT-coated glass substrates, and afterwards the functionalizations
were burnt to obtain bare CNTs deposited on the glass. CNTs are capable to boost
neuronal network activity6 but the mechanism is still unclear. To address this
issue, we produced glass substrates covered with CNTs and studied the
morphology and the interactions of the neuronal network. Dual
electrophysiological recordings were also performed proving that the connectivity
in the neuronal network is improved when CNTs are present.
References
(1) Prato, M.; Kostarelos, K.; Bianco, A. Functionalized carbon nanotubes in drug
design and discovery. Accounts of Chemical Research 2008, 41, 60-68.
(2) Wu, W.; Wieckowski, S.; Pastorin, G.; Benincasa, M.; Klumpp, C.; Briand, J.;
Gennaro, R.; Prato, M.; Bianco, A. Targeted delivery of amphotericin B to cells
by using functionalized carbon nanotubes. Angewandte Chemie - International
Edition 2005, 44, 6358-6362.
(3) Guryanov I.; Toma F. M.; Montellano López A.; M., C.; Da Ros T.; Angelini G.;
D'Aurizio E.; Fontana A.; Prato M.; Bonchio M. Microwave-Assisted
Functionalization of Carbon Nanostructures in Ionic Liquids. Chemistry - A
European Journal.
(4) Herrero, M. A.; Toma, F. M.; Al-Jamal, K. T.; Kostarelos, K.; Bianco, A.; Da Ros,
T.; Bano, F.; Casalis, L.; Scoles, G.; Prato, M. Synthesis and Characterization of
a Carbon Nanotube−Dendron Series for Efficient siRNA Delivery. Journal of
the American Chemical Society 2009, 131, 9843-9848.
(5) Gao, L.; Nie, L.; Wang, T.; Qin, Y.; Guo, Z.; Yang, D.; Yan, X. Carbon nanotube
delivery of the GFP gene into mammalian cells. Chembiochem 2006, 7, 239-
Thesis abstract
7
242.
(6) Lovat, V.; Pantarotto, D.; Lagostena, L.; Cacciari, B.; Grandolfo, M.; Righi, M.;
Spalluto, G.; Prato, M.; Ballerini, L. Carbon Nanotube Substrates Boost
Neuronal Electrical Signaling. Nano Letters 2005, 5, 1107-1110.
Thesis abstract
8
Chapter 1
9
Chapter 1:
General INTRODUCTION
And
Aim of the thesis
1.1 Nanotechnology: Small Can Be Big!
In spite of few extremely visionary sentences about the fact that at the
bottom of size scale “an enormous amount (of work) can be done in principle” and
“that it would have an enormous number of technical applications” [Feynman in
his famous 1959 speech], nobody never talked about “Nanotechnology” before
1974, when this term was introduced by N. Taniguchi1,2. Moreover, in the 1980s E.
Drexler explored the possibility to build machines manipulating individual
atoms2,3, ending up with a book (the first Nanotechnology book) written in 19864.
On the other hand, this obviously does not mean that nanostructures did not exist
before that time. They are as old as life itself, since humans soon took advantage of
nanosized materials. For example, the first archeological find, the Lycurgus cup
(British Museum, London, Figure 1) dated from the 4th century A.D., contains silver
and gold nanoparticles5.
Chapter 1
10
Figure 1. Lycurgus cup, IV A.D., Roman find. A: normal view of the cup; B: view of the cup when
illuminated from the inside. It illustrates King Lycurgus’ death. © Trustees of the British Museum.
Nanotechnology concerns all structures of nanometer (a billionth of a
meter) size, even though also microstructures are often considered as a part of
Nanotechnology because, unless they are made of a single kind of atoms, they often
contain nanoscale sub-structures.
One of the first implications in this field is the dominant role of the surface.
Consider a cube of 1 cm side, the volume is 1 cm3 and the surface area is 6 cm2; in
contrast, if we subdivide the cubic centimeter into 1021 cubes of 1 nm size, the total
volume will remain the same (1 cm3) but the surface area will be 6x1021 nm2 (i.e.
6000 m2), that is more or less equivalent to the surface of a soccer field. It appears
clear that small can be really big! The increase of surface area is very useful in
many fields as, for example, in drug delivery or catalysis. This is only one of the
many features of the nano-range, indeed nanostructures possess many other
peculiar properties, originating a wide range of applications in physics, chemistry,
engineering, biology and medicine. As a matter of fact several new sub-disciplines
have been recently born such as, for example, “nanomedicine” or “nanometrology”.
A
B
Chapter 1
11
The discovery and the subsequent applications of CNTs have allowed the
development of an entire branch of nanotechnology based on this versatile
material5, the material with which all experiments reported in this thesis have
been carried out.
1.2 Carbon Nanotubes6
Carbon is the only element of the periodic table that occurs in allotropic
forms from 0 dimensions to 3 dimensions, due to its different hybridization
capabilities. Hybridization occurs when electronic wave functions of valence
electrons can mix with each other and change the occupation of the outer shell s
and p orbitals (2s and 2p for the carbon atom). The mixing of a 2s orbital with n= 1,
2, 3 2p electrons is called spn hybridization. Fullerene (zero-dimensional), carbon
nanotubes (one-dimensional) and graphene (single layer of graphite, two-
dimensional) are all made of sp2 hybridized carbon atoms, whereas diamond
(three-dimensional) are sp3 hybridized; amorphous carbon (three-dimensional)
lays in between graphite and diamond with a mixed sp2 and sp3 hybridization.
Graphite is a three dimensional layered hexagonal solid lattice of carbon
atoms, one single layer of which is a two-dimensions material namely graphene. It
is important to introduce graphite because many properties of carbon nanotubes
are the direct consequence of graphite properties. To illustrate this point, in Figure
2 we show the unit cell of graphene, where 𝑎1 and 𝑎2 are unit vectors, in Cartesian
coordinates (1).
Chapter 1
12
Figure 2. Unit cell of graphite
𝑎1 = 3
2 a,
a
2 𝑎2 =
3
2 a, −
a
2 (1)
Where (a) is the lattice constant of two dimensional graphite (2).
𝑎 = 𝑎1 = 𝑎2 = 1.42 × 3 = 2.46 Å (2)
Carbon Nanotubes were discovered by Iijima in 19917, who was awarded in
2008 of the Kavli prize. CNTs can be imagined as a graphene sheet rolled up into a
tube, and this is why many of CNTs properties are comparable with those of
graphite. Mainly two different types of CNTs can be distinguished: (i) single walled
CNTs (SWCNTs) and (ii) multi walled CNTs (MWCNTs). They are generally made of
a side wall and two end caps: the sidewall is made of benzene rings and end caps
are fullerene-like.
- SWCNTs are composed of a single layer of graphene and their diameter
ranges between 0.7 and 1.4 nm depending on the temperature they have been
synthesized. In fact, it has been observed that a higher growth temperature gives a
larger diameter. This is probably because when the temperature is diminished the
probability of generating a pentagonal ring is increased and the generated caps
tend to be smaller. At higher temperature, vibrational entropy favors hexagonal
rings over pentagonal rings and consequently the diameter is larger. SWCNTs can
Chapter 1
13
be considered as one dimensional object because of their large aspect ratio (i.e.
length/diameter = 104-105). Another essential fact about their structure is related
to the six-membered carbon ring orientation in the honeycomb lattice relative to
the axis of SWCNTs, that determines a parameter known as chirality (𝐶ℎ , Figure 3).
Figure 3. A: unrolled honeycomb lattice of a SWCNT, 𝑂𝐴 , 𝑂𝐵 and 𝑂𝑅 are the chiral vector (𝐶ℎ ),
translational vector (𝑇 ) and the symmetry vector (𝑅 )respectively. B, C, D:classification of SWCNTs,
armchair, zigzag and chiral respectively. Armchair and zigzag refer to the shape of the cross sectional
ring. is the chiral angle and denotes the tilt angle of the hexagons with respect to the direction of
nanotube axis. Adapted from reference8.
Fundamentally, the primary classification of SWCNTs is done on the symmetry
basis: a SWCNT can be achiral (armchair or zigzag) or chiral, in the latter case the
mirror image is not identical to the original structure because the CNT is a spiral.
As reported in Figure 2, also in Figure 3 we can distinguish unit vectors of
hexagonal lattice (𝑎1 and 𝑎2 ) that define chirality (3):
𝐶ℎ = 𝑛𝑎1 +𝑚𝑎2 (3)
n and m are integers and on this basis we can distinguish three different situations:
n = m is the case of armchair structure, m = 0 is the case of zigzag structure and all
other cases describe chiral nanotube structures. Chirality is important to the
defined conductivity: armchair carbon nanotubes are metallic, zigzag and chiral
A
B
C
D
Chapter 1
14
CNTs are semiconducting. 𝑂𝐵 , which is the translation vector (𝑇 ), is parallel to the
nanotube axis and normal to 𝐶ℎ (Figure 2). 𝑇 corresponds to the first lattice point
of the two-dimension graphene sheet through which this vector passes. We need
one more vector to denote SWCNTs structure: this is the symmetry vector (𝑅 ),
used for generating the coordinates of carbon atom in the nanotube itself.
Transmission electron microscopy (TEM, Appendix A) and above all scanning
tunneling microscopy (STM), with results reported also in our group9
experimentally confirmed this structure (Figure 4).
Figure 4. A: TEM image of HiPCO SWCNT (scale bar 100 nm). B: STM image of HiPCO SWCNT9.
SWCNTs are usually organized into bundles or ropes (see Figure 3A), and it is
difficult to proceed in their exfoliation. - stacking occurs between the sidewalls
and this is why they stack one next to the other. Moreover a hydrophobicity
contribution is also expected in determining this entanglement6. The order of these
intermolecular forces is about 0.5 eV/nm10.
- MWCNTs are made of many graphene layers and the diameter can be up to
100 nm.
Two main models describe MWCNTs: (i) the Russian doll model or coaxial
cylindrically curved model in which many graphene layers are rolled up and
arranged in a cylindrical fashion, and (ii) the parchment model in which a single
A
B
Chapter 1
15
graphene layer is scrolled as a parchment (Figure 5a and c respectively). A further
evolution of the coaxial cylindrically curved model is the coaxial cylindrically
poligonized (Figure 5b).
Figure 5. Different models of MWCNTs. (a) coaxial cylindrically curved; (b) coaxial poligonized; (c) scroll
graphene sheet8.
The electronic structure of a MWCNT is considered as the sum of the electronic
structures of the constituent nanotubes. Thereby, we state that in the limit of the
large nanotube diameter, all the nanotubes properties correspond to those of
graphite. The interlayer distance in MWCNT, in fact, is equal to the interlayer
distance in graphite (3.35 Å)6. These interactions are not sufficient to correlate the
chirality of near tubes and MWCNTs are a mix of achiral shells. The result is that
the lattice structures of the inner and outer layers are generally incommensurate.
Physical properties of MWCNTs have not been explored in detail, because of the
difficulty in performing measurements on each single layer. In figure 6, TEM (A)
and scanning electron microscopy (SEM) (B) images of MWCNTs are reported. The
number of inner shells can be measured with High Resolution TEM.
Chapter 1
16
Figure 6. A: TEM image of double walled CNTs (scale bar = 100nm); B: SEM image of MWCNTs (scale bar
= 1m).
1.2.1 Synthesis and Functionalization of CNTs
There are several methods to synthetize CNTs:
Arc discharge is actually the most common procedure to produce carbon
nanostructures. It consists of two electrodes of graphite and a DC voltage is
applied between them. The whole system is under pressure control with
both a vacuum and a gas line, and MWCNTs are the main product, with
fullerenes and amorphous carbon particles as by-products. SWCNTs can be
produced as well, varying pressure and temperature conditions. Nanotube
are producted at the cathode, while at the anode there is a hole filled with
Ni-Co, Co-Y or Ni-Y, a source of metal nanoparticles, catalysts, those consist
an impurity of CNTs11.
Laser ablation methods were developed by R. E. Smalley (Nobel Laureate in
1996 with R. F. Curl and H. W. Kroto for the discovery of fullerene) at the
Rice University. The apparatus consists of a tube, where Ar or He is fluxed,
within a furnace. Inside the tube there is a target of graphite mixed with 1%
of Co and 1% of Ni, which is vaporized using a pulsed laser and high
temperature. Carbon nanostructures are collected in an external trap, while
B
A
Chapter 1
17
metal nanoparticles are not trapped inside the cage. Both SWCNTs and
MWCNTs can be obtained with this technique11.
Chemical Vapor Deposition (CVD) is one of the main interesting synthetic
procedures because it employs hydrocarbon decomposition over
substrates, where metal nanoparticles have been placed. Ni, Fe or Co, or
alloys of these three elements, are used. Impurities due to the pyrolytic
graphite are few, but metal nanoparticles are presents, especially at the
periphery of CNTs. MWCNTs are grown by decomposition of acetylene.
Further, SWCNTs are produced by floating CO with Fe(CO)5 with a
disproportionation process, known as “HiPCO” (High Pressure CO)11.
Beyond the different techniques, it appears immediately clear that CNTs need
processing after their synthesis. To address this issue, purification methods and,
above all, functionalization approaches are essential to allow manipulation and
further application of this material. Usually, metal nanoparticles and amorphous
carbon are present as a synthetic residue. In general, CNTs are a fluffy powder
difficult to manage, while chemical functionalization contributes to prepare more
homogenous and soluble material (Figure 4).
Figure 7. (A) Powder of SWCNTs, (B) DMF solution of f-SWCNTs (via 1,3-dipolar cycloaddition, SWCNTs 9
Chapter 2, Table 2).
Purification concerns acidic treatments that bring to the removal of
amorphous particles and metals oxidation. Moreover, aggressive purification
procedure can shorten CNTs and produce oxidized nanotubes.
B
B
A
Chapter 1
18
On the other hand, a wide variety of functionalizations have been reported
in the literature10,12,13, the most important of which are summarized in Scheme 1.
Scheme 1. Most common functionalizations of CNTs are reported. Adapted from reference12.
We can categorize mainly non-covalent and covalent approaches.
Non-covalent functionalization is based on van der Waals, hydrophobic or
- interactions (Scheme 1), and it is particularly interesting because it does not
perturb the electronic structure of CNTs, and of SWCNTs in particular. This
functionalization involves weak forces, and so some applications are prevented.
The systems, in fact, are not only difficult to control but also difficult to
characterize. The amount of weakly bound molecules is not always calculable;
moreover, especially in solution, other interactions can occur and these molecules
can be replaced by solvent. Anyway, many examples of non covalent interaction
are reported: surfactants are widely used to exfoliate bundles of SWCNTs14, as well
as ionic liquids (discussed in Chapter 2) in order to facilitate CNTs manipulation
and further reactions. CNTs wrapping by polymers, including DNA, have been
studied13; also proteins are able to non convalently interact with CNTs and these
are often used for bio-sensor applications11.
Chapter 1
19
Endohedral is a non convalent functionalization that occurs trapping molecules
inside the CNT structure13.
In covalent functionalization, CNTs are functionalized by not reversible
attachment of appendage on the sidewalls and/or on the tips. Also in this case,
many different approaches are reported13. Briefly, reactions can be performed at
the sidewall site (sidewall functionalization) or at the defect sites (defect
functionalization), usually localized on the tips. In the first case, fluorination with
elemental fluorine at high temperature (400-600°C) has been explored,
accomplishing further substitutions with alkyl groups. Furthermore, radical
addition via diazonium salt has been proposed by Tour’s group. On the other hand,
cycloadditions have found wide interests. Besides 1,3-dipolar cycloadditions of
azomethine ylides that is described in Chapter 2 and Appendix B, other
cycloadditions have been reported, such as carbene [2+1] cycloadditions or Diels-
Alder via microwave (MW) irradiation cycloaddition13. Side defects
functionalization occurs via amidation or esterification reactions of carboxylic
residues obtained on CNTs. Moreover, it is feasible that in general caps (i.e. tips
when they are not cut) are more reactive than sidewalls because of their mixed
pentagonal-hexagonal structure.
f-MWCNTs are not modified in their electronic structure and new
properties can be added by means of functionalization. Instead, electronic
properties of SWCNTs are perturbed by covalent functionalization and double
bonds are irreversibly lost. This may affect conductive property, preventing
further CNT applications.
1.2.2 Applications and Properties of CNTs11
Mechanical, thermal and conductive properties of carbon nanotubes are
outstanding.
For what concerns mechanical properties, the Young modulus, that gives
information about the stiffness of material, has been calculated for MWCNT and
Chapter 1
20
SWCNT and, depending on the different type of CNT, it varies from 0.5 up to 1.1
TPa. For comparison, the Young modulus of diamond is 1.2 TPa. Moreover, it has
been demonstrated that the breaking strength of a CNT, attached to two AFM tips,
is 11-63 GPa for a MWCNT (referred to the outer shell), and 30 GPa (up to 50 GPa)
for a SWCNT. These values proved that the mechanical properties of CNTs are very
good: in fact, they are related to the cylindrical structure of CNTs that confers
structural stability under compression.
Thermal conductivity seems to be higher than for every other material and
experiments under ultra high vacuum showed that CNTs are stable at least at 4000
K. In fact, carbon nanotubes benefit of both sp2 and one-dimensional structure.
Conductive properties are related with the one-dimensional structure of
CNTs as well and electrons undergo ballistic conduction. This type of conduction is
defined as continuous flow of energy carrying over high distances without
stoppers.
Interesting optical properties have been studied in the last few years and
many advances in photophysics have been achieved.
The power of CNTs resides on the vast number of applications of this
versatile material, which go from electronics to nanomedicine. The final challenge
remain in achieving large-scale production and industrial applications of CNTs.
o Carbon nanotubes applications to electronic device fabrication are due to
their superlative conductive properties. Transistors, such as field-effect transistors
(FETs), have been successfully realized; moreover attempts to develop
electromechanical memory devices are under study. CNTs-FETs are used in gas
sensing: sensitivity and selectivity of these transistors are improved using f-CNTs,
attached to binding ligands. No commercial unit built in this way is available, as far
as we know. Instead, thin film of SWCNTs have been incorporated also in
transparent transistors for military or industrial electronic systems. Among
electronic applications, we can count also CNTs as field emitters. It basically means
that CNTs own the peculiar property to emit electrons when an electromagnetic
field is applied, because of their peculiar structure (electric field are much stronger
Chapter 1
21
at the tips). While, concerning their thermal properties, aligned CNTs on silicon
chips provide high cooling performance for microprocessors.
o Energy applications use both CNTs as electrode material or as conductive
filler for the active material. Electronic and electrical properties together with the
high surface area are exploited. Energy storage devices such as lithium ion battery
and supercapacitors have been investigated, as well as solar cells. In the latter case,
silicon-based photovoltaics with solar-energy conversion of 25% and CNTs
forming bulk heterojunctions in solar organic cells have been built. Electrodes of
CNTs are employed also in electrochemical sensors due to faster electron-transfer
kinetics from a number of electroactive-species. These electrodes are often
exploited as biosensors for glucose or DNA hybridization.
o Mechanical applications of CNTs are directed to obtain CNTs based
composites: NASA is investing large amounts of resources in developing these
materials for applications in aerospace.
o There are many examples of biological applications in literature. In this
case, CNTs must be water-soluble to avoid cytotoxicity. Thus, functionalization is
necessary and covalent functionalization often represents the first choice. CNTs
have been exploited for imaging but also as drug vehicles or carriers. This is
possible because CNTs are readily internalized by cells and it is possible to
introduce a multiple functionalization with different therapeutic agents. As cargo
system, they have been developed to carry DNA, proteins or anticancer drugs.
However, internalization and excretion and interferences at a systemic level (i.e
metabolism) still remain unclear. It seems that different types of functionalization
could bring different fate to the CNTs: Dai et al15 proposed an endocytosis pathway
for acid-functionalized SWCNTs, while Bianco et al16 suggested a passive
mechanism in which nanotubes could enter cell like needles. Besides this, it is
undeniable that functionalized, water-soluble, CNTs bear an extraordinary affinity
to penetrate cells. One of the factors that influences CNT-cell interaction could be
also represented by protein adsorption at nanotube surface. Also receptor-
mediated mechanism have been proposed17. On the other hand, another tricky
Chapter 1
22
question is the destiny of CNTs once inside the cell. Interestingly, Strano and
collaborators suggested the first evidence of exocytosis of a SWCNT, reporting that
the exocytosis and endocytosis have almost the same rate18. At systemic level, it
seems that CNTs are excreted by kidneys19. We are still pursuing these issues and
many different hypothesis and mechanisms, also proposed in this thesis (Chapter
3), concerning functionalizations, drug delivery and cellular uptake and CNTs
methabolism are inserting themselves in the ongoing discussions on these
subjects.
1.3 CNT toxicity
Toxicity of CNTs represents a hot topic. It relates to the health care of
researchers working with this nanomaterial, but it is also an issue concerning all
their future applications. In fact, this general anxiety is about more or less all
nanomaterial, and nanotoxicology is a new science that verifies the acute and
chronic toxicity of all new nanoscale materials. Another problem, to be mentioned,
is that nanotechnology and carbon nanotubes have been studied for approximately
twenty years and worries about their toxicity seriously arised during the last ten
years. This means that results about long-term toxicity are still far to be defined.
Anyway, different studies with as-produced and f-SWCNTs and f-MWCNTs
have been performed, sorting out that the toxicity is related to purity and
functionalization of CNTs. Wick et al have shown that on lung mesothelic cells
(MSTO-211H) cytotoxicity is correlated to the degree of agglomeration. They
tested SWCNTs as produced and found that suspended CNTs-bundles are less toxic
than asbestos fibers20. Nevertheless, it is common opinion that chemically f-CNTs
do not exert cytotoxicity21,22, and, in particular and not surprisingly, that covalent
f-CNTs are better than non-covalent functionlized ones, depending on the degree of
functionalization23.
Besides this, a paper appeared on Nature Nanotechnology24 had an impact
in that field that in our opinion has been excessive. These authors reported that
Chapter 1
23
CNTs show asbestos-like pathogeniticity, but their pilot study was developed using
long MWCNTs (from 5 to 20 m). In our study, and in general for biological
applications, CNTs are shorter than 1 m and the average length is usually
between 400-600 nm. Moreover, we can add that, for what concerns asbestos, fiber
length and diameter are important factors in determining the citotoxicity25 and it
seems that longer fibers are more cytotoxic than shorter ones, even though
authors claimed that the debate was still open. To our knowledge, shorter CNTs
are also more soluble and then biocompatible.
The toxicological properties of all nanomaterials are not negligible, but a lot
of studies are still necessary. From what is reported above, it is clear that covalent
functionalization improves the biocompatibility and that the length of CNTs has to
be controlled. Indeed, in this thesis, we demonstrate the full biocompatibility of
short covalently f-CNTs.
1.4 Aim of the thesis
The uses of CNTs in different fields are really promising and the aim of this work is
the preparation of new functionalized CNTs with different applications, focusing
our attention on biological purposes.
One of the main limitations of CNTs is the lack of solubility in solvents, especially
the biological compatible ones. To overcome this problem, we covalently
funtionalize carbon nanotubes using two different approaches. The first one is
represented by 1,3-dipolar cycloaddition of azomethine ylides by means of thermal
heating, the second one concerns the use of MW irradiation combined with the
presence of ionic liquids (ILs) to functionalize CNTs with differently substituted
pyrrolidine rings (Chapter 2). On the surface of the derivatives obtained in the
thermal conditions we will grow dendron units to increase the water-solubility.
Different dendron generations will be prepared to study their solubility, their
cellular uptake and their capability to act as siRNA delivery systems (Chapter 3).
We will use a RNA sequence able to kill cells, so the effects of the delivery will be
Chapter 1
24
unequivocal. Moreover we will use these compounds to perform preliminary
biodistribution studies, necessary to develop the use of CNT for biological
applications.
The CNT derivatives will be also explored as substrates for neuronal
cultures (Chapter 4). Morphological studies to investigate the interactions between
CNTs and cellular membranes will be performed and electrophysiological
recordings will be used to characterize the neuronal network.
References
(1) Taniguchi, Norio On the Basic Concept of Nanotechnology. Proc. ICPE Tokyo
1974, Part II, 18-23.
(2) Edwards, S. A. The Nanotech Pioneers: Where Are They Taking Us; Wiley-VCH,
2006.
(3) Drexler, K. E. Molecular engineering: An approach to the development of
general capabilities for molecular manipulation. Proceedings of the National
Academy of Sciences of the United States of America 1981, 78, 5275-5278.
(4) Drexler, E. Engines of Creation: The Coming Era of Nanotechnology; Anchor,
1987.
(5) Jr, C. P. P.; Owens, F. J. Introduction to Nanotechnology; 1° ed.; Wiley-
Interscience, 2003.
(6) Saito, R.; Dresselhaus, G.; Dresselhaus, M. S. Physical Properties of Carbon
Nanotubes; 1° ed.; World Scientific Publishing Company, 1998.
(7) Iijima, S. Helical microtubules of graphitic carbon. Nature 1991, 354, 56-58.
(8) Belin, T.; Epron, F. Characterization methods of carbon nanotubes: A review.
Materials Science and Engineering B: Solid-State Materials for Advanced
Technology 2005, 119, 105-118.
(9) Bonifazi, D.; Nacci, C.; Marega, R.; Campidelli, S.; Ceballos, G.; Modesti, S.;
Meneghetti, M.; Prato, M. Microscopic and spectroscopic characterization of
paintbrush-like single-walled carbon nanotubes. Nano Letters 2006, 6, 1408-
1414.
Chapter 1
25
(10) Singh, P.; Campidelli, S.; Giordani, S.; Bonifazi, D.; Bianco, A.; Prato, M. Organic
functionalisation and characterisation of single-walled carbon nanotubes.
Chemical Society Reviews 2009, 38, 2214-2230.
(11) Jorio, A.; Dresselhaus, Gene; Dresselhaus, Mildred S. Carbon Nanotubes:
Advanced Topics in the Synthesis, Structure, Properties and Applications; 1° ed.;
Springer, 2008.
(12) Hirsch, A. Functionalization of single-walled carbon nanotubes. Angewandte
Chemie - International Edition 2002, 41, 1853-1859.
(13) Tasis, D.; Tagmatarchis, N.; Bianco, A.; Prato, M. Chemistry of carbon
nanotubes. Chemical Reviews 2006, 106, 1105-1136.
(14) Vaisman, L.; Wagner, H.; Marom, G. The role of surfactants in dispersion of
carbon nanotubes. Advances in Colloid and Interface Science 2006, 128-130,
37-46.
(15) Kam, N.; Jessop, T.; Wender, P.; Dai, H. Nanotube molecular transporters:
Internalization of carbon nanotube-protein conjugates into mammalian cells.
Journal of the American Chemical Society 2004, 126, 6850-6851.
(16) Pantarotto, D.; Singh, R.; McCarthy, D.; Erhardt, M.; Briand, J.; Prato, M.;
Kostarelos, K.; Bianco, A. Functionalized carbon nanotubes for plasmid DNA
gene delivery. Angewandte Chemie - International Edition 2004, 43, 5242-
5246.
(17) Kam, N.; Liu, Z.; Dai, H. Carbon nanotubes as intracellular transporters for
proteins and DNA: An investigation of the uptake mechanism and pathway.
Angewandte Chemie - International Edition 2006, 45, 577-581.
(18) Jin, H.; Heller, D.; Strano, M. Single-particle tracking of endocytosis and
exocytosis of single-walled carbon nanotubes in NIH-3T3 cells. Nano Letters
2008, 8, 1577-1585.
(19) Lacerda, L.; Herrero, M.; Venner, K.; Bianco, A.; Prato, M.; Kostarelos, K.
Carbon-nanotube shape and individualization critical for renal excretion.
Small 2008, 4, 1130-1132.
Chapter 1
26
(20) Wick, P.; Manser, P.; Limbach, L.; Dettlaff-Weglikowska, U.; Krumeich, F.;
Roth, S.; Stark, W.; Bruinink, A. The degree and kind of agglomeration affect
carbon nanotube cytotoxicity. Toxicology Letters 2007, 168, 121-131.
(21) Smart, S.; Cassady, A.; Lu, G.; Martin, D. The biocompatibility of carbon
nanotubes. Carbon 2006, 44, 1034-1047.
(22) Dumortier, H.; Lacotte, S.; Pastorin, G.; Marega, R.; Wu, W.; Bonifazi, D.;
Briand, J.; Prato, M.; Muller, S.; Bianco, A. Functionalized carbon nanotubes
are non-cytotoxic and preserve the functionality of primary immune cells.
Nano Letters 2006, 6, 1522-1528.
(23) Sayes, C.; Liang, F.; Hudson, J.; Mendez, J.; Guo, W.; Beach, J.; Moore, V.; Doyle,
C.; West, J.; Billups, W.; Ausman, K.; Colvin, V. Functionalization density
dependence of single-walled carbon nanotubes cytotoxicity in vitro.
Toxicology Letters 2006, 161, 135-142.
(24) Poland, C.; Duffin, R.; Kinloch, I.; Maynard, A.; Wallace, W.; Seaton, A.; Stone,
V.; Brown, S.; MacNee, W.; Donaldson, K. Carbon nanotubes introduced into
the abdominal cavity of mice show asbestos-like pathogenicity in a pilot
study. Nature Nanotechnology 2008, 3, 423-428.
(25) Goodglick, L.; Kane, A. Cytotoxicity of long and short crocidolite asbestos
fibers in vitro and in vivo. Cancer Research 1990, 50, 5153-5163.
Chapter 2
27
Chapter 2:
Covalent
Functionalization of
Carbon Nanotubes
2.1 Synthesis and Physicochemical characterization of a Carbon
Nanotube−Dendron Series1
By means of widely used thermal reactions we succeeded in the synthesis of
a new series of polyamidoamine (PAMAM) dendron-functionalized MWCNT
derivatives, characterized by the presence of numerous tetra-alkyl ammonium
salts at the periphery of the dendron.
The polyamidoamine (PAMAM) dendrons attached onto MWCNTs1 enhance
their aqueous solubility due to the presence of many polar groups. A stepwise
synthetic process was followed to achieve growth of the dendron on the MWCNTs
1 The work reported in the first part of this chapter has been published in: Journal of American
Chemical Society, 2009, 131 (28), pp 9843-9848. “Synthesis and characterization of a carbon
nanotube-dendron series for efficient siRNA delivery” by Herrero M.A., Toma F.M., Al-Jamal K.T.,
Kostarelos K., Bianco A., Da Ros T., Bano F., Casalis L., Scoles G., Prato M.
Chapter 2
28
up to the second-generation. Pristine MWCNTs, initially functionalized using 1,3-
dipolar cycloaddition of azomethine ylides, were further modified by combining a
double additions of methyl acrylate, including a complete addition of the amino
groups, with ethylene diamine addition, leading to an exhaustive amidation of the
ester functionalities (Scheme 1). The second generation dendron bears the highest
number of peripheral primary amino groups (4 amino groups per dendron
moiety), which eventually undergo a process of alkylation using an epoxy
derivative2. The same alkylation was also performed for zero, first and second-
generation (G0, G1 and G2 respectively).
In detail, functionalization of MWCNTs using the 1,3-dipolar cycloaddition
of azomethine ylides was used to introduce the pyrrolidine moiety at the external
surface and the tips of MWCNTs 1 (Scheme 1). First-generation (G1) and second-
generation (G2) dendron-MWCNT 4 and 7 were synthesized by repeating twice the
two-step reaction using methyl acrylate and ethylene diamine. We proceeded in
alkylation of the (G0) dendron-MWCNT 1, (G1) dendron-MWCNT 4 and (G2)
dendron-MWCNT 7 with a glycidyl trimethylammonium chloride epoxy derivative
to generate polycationic dendron-MWCNT 2, 5 and 8, respectively (Scheme 1)2.
The Kaiser test allowed quantification of the amino groups for the different
generations of dendrons before alkylation3. The test gave negative results for the
alkylated conjugates (2, 5 and 8), consistent with complete conversion of the
primary amino groups. The amount of primary amino groups on MWCNT 1 was
measured by the Kaiser test as 481 mol/g. Subsequent Kaiser tests gave amine
loadings of 718 mol/g and 1097 mol/g for the first and second generation,
respectively.
Chapter 2
29
Chapter 2
30
Scheme 1. Reaction conditions for the synthesis of G2 PAMAM dendron-MWCNTs and consequent
alkylation of primary amino groups with glycidyl trimethylammonium chloride. MeOH, 40 °C, 48 h. R = -
CH2CH(OH)CH2N(CH3)3+Cl-.
The different CNT conjugates were fully characterized by standard techniques, as
thermogravimetric analysis (TGA), TEM, and both atomic force microscopy (AFM)
and enhanced electrostatic force microscopy (E-EFM).
The solubility of the charged dendron-MWCNTs was assessed in water. The
insets of Figure 1 display the differences in the dispersibility of the conjugates. A
clear increase in solubility was observed going from MWCNT 1 to the higher
alkylated dendron generations. The aqueous solution of compound 8 was stable
for at least four weeks at room temperature without formation of any precipitate
and the solubility obtained was up to 3 mg/mL.
Chapter 2
31
Figure 1. TEM image of (A) MWCNT 1; (B) alkylated (G0) dendron-MWCNT 2; (C) alkylated (G1) dendron-
MWCNT 5; (D) alkylated (G2) dendron-MWCNT 8. Scale bar is 1 m. Inset: MWCNT 1, alkylated (G0)
dedron-MWCNT 2, alkylated (G1) dendron-MWCNT 5, and alkylated (G2) dendron-MWCNT 8 (all
solubilized at 1 mg/mL of H2O).
A B
C D
Chapter 2
32
Figure 2. A) (G2) dendron-MWCNT 7 (0.3 mg) in 3 mL of DMF and H2O, respectively; B) alkylated (G2)
dendron-MWCNT 8 (0.3 mg) in 3 mL of DMF and H2O, respectively. C) TEM image of (G1) dendron-
MWCNT 4 (dissolved in water), and D) TEM image of alkylated (G2) dendron-MWCNT 8 (dissolved in
water). Scale bar is 1 m.
Moreover, the water solubility of conjugate 8 was tenfold higher than
conjugate 7 with solubility of 3 mg/mL and 0.1 mg/mL, respectively. Figure 2A
shows that dendron-MWCNT 7 disperses better in dimethyl formammide (DMF),
with relatively low solubility in water. Alkylated conjugate 8, on the other hand,
exhibits reversed behavior, with enhanced solubility in water (Figure 2B). TEM
images confirm the presence of MWCNTs, showing morphological differences for
the different dendron generations in water, including a remarkable increase in the
degree of dispersibility (Figure 2C and 2D).
Chapter 2
33
Figure 3. TEM image of (A) Boc-amino protected MWCNTs 1, and (B) MWCNT 6. Scale bar: 200 nm.
Figure 4. TEM images of dendron-MWCNT 6. Scale bar: 200 nm (A) and 100 nm (B).
The f-MWCNTs were structurally studied by TEM (Figure 1-4). The images confirm
the presence of MWCNTs, showing morphological differences for the different
dendron generations in water, including a remarkable increase in the degree of
dispersibility.
In addition to TEM, alkylated (G2) dendron-MWCNT 8 was characterized by
AFM, performed by depositing a sample of the aqueous dispersion on a mica
surface. A progressive increase in the diameter of the nanotubes as the dendritic
structure grew was observed. The AFM images show in the sample both
individualized and small bundles of MWCNTs. The 3D view of cropped section of
Boc-amino protected MWCNT 1 and dendron-MWCNT 8 from a 1.2 × 1.2 μm2 AFM
topographic image along their line profiles are shown in Figure 5A and 5B,
A B
A B
Chapter 2
34
respectively. As can be seen, there is a clear increase in diameter of the MWCNT
conjugates with the increasing of dendron generations. To evaluate the average
profile, the diameter of pristine MWCNTs, Boc-amino protected MWCNT 1 and
dendron-MWCNT 8 was measured, giving values that range from 21-25 nm for
Boc-amino protected MWCNT 1 and up to 36-40 nm for MWCNT 8, respectively
(Figure 5C).
Figure 5. 1.2 × 1.2 μm2 AFM topographic images of (A) MWCNTs after 1,3-dipolar cycloaddition reaction
(Boc-amino protected MWCNT 1), and (B) the alkylated (G2) dendron-MWCNTs 8 along their 3D view of
Chapter 2
35
cropped segments [red square box in part (A) and (B) and line profiles (right plots)]. (C) Diameter
distribution of pristine MWCNTs (black), MWCNT after 1,3-dipolar cycloaddition reaction (Boc-amino
protected MWCNT 1) (red), and alkylated (G2) dendron-MWCNT 8 (blue). The three curves represent the
Gaussian fit of the data.
The weight loss for the highly functionalized conjugates, calculated from the
thermogravimetric curves at 450°C, was directly correlated to the increase of mass
around the CNTs, introduced at each step. The mass attributed to functionalities
increased from 12.5% in the first-generation 4 to 18 % in the second-generation
dendron-MWCNT 7. The final derivative 8 showed a 28% mass loss (Figure 6). The
number of the functional groups for the final dendron-MWCNT 8 calculated from
the TGA curve corresponds to 237 mol/g, resulting in 948 mol/g of amine
groups for the precursor MWCNT 7. This value was in good agreement with the
value found using the Kaiser test (1097 mol/g).
Chapter 2
36
Figure 6. TGA curves of pristine MWCNTs (black dash dot line), G1 dendron-MWCNT 4 (blue dot line), G2
dendron-MWCNTs 7 (red solid line) and alkylated G2 dendron-MWCNT 8 (green dash line). All the
experiments were performed under N2 atmosphere.
In order to assess the presence of functionalized groups on the MWCNT
surface, we investigated the electrostatic properties of dendron-functionalized
MWCNTs deposited on a silicon substrate. By means of E-EFM we studied the
localization of positive charges due to the quaternary salt of the moieties.
We investigated, at different applied biases, two different samples: pristine
MWCNTs and MWCNTs 8, each of them deposited as methanolic solutions on a
silicon oxide surface.
Figure 7. E-EFM phase images respectively at a) -3V, b) 0V and c) 3V, of pristine MWCNTs; d: topography
of pristine MWCNTs.
Chapter 2
37
Figure 8. Phase shift versus bias plot of: background (green square symbol), nanotube body (blue
triangular symbol) and “sweelings” (red circular symbol) for pristine MWCNTs on silicon oxide. The
phase-shift is related to the different regimes: repulsive for negative phase shift, and attractive for
positive phase shifts.
We reported the topography, the simultaneosly acquired phase images and
corresponding plots of pristine MWCNTs. Herein, we noticed that some swellings,
as impurities, appeared also in the case of not-functionalized nanotubes.
Noteworthy, impurities have a different phase shift with respect to the
functionalizations of MWCNTs. In fact, we plotted E-EFM phase signals for three
different areas for unmodified MWCNTs:silicon oxide background, nanotubes and
“swellings”. The phase shift versus applied probe bias is shown in Figure 7a, b and
c. We can assert that the total charge on the surface is negative and this is proved
by graph trend in Figure 7 (repulsion for negative biases, attraction in case of
positive voltages). This general behavior may be explained by the presence of
negative terminations distributed on the silicon oxide surface. The behavior of the
Chapter 2
38
swellings, in this case, does not have any difference compared to that one of the
nanotubes.
We proceeded in the analysis of MWCNT 8. In the topographic image shown
in Figure 9D, we can distinguish three different areas: a homogeneous substrate
(i), the nanotubes, uniformly dispersed on the substrate (ii), and some
“decorations” (iii) that appear along the nanotubes in the case of dendron f-
MWCNTs.
Figure 9. E-EFM phase images respectively at A) -3V, B) 0V and C) 3V, of derivative dendron f-MWCNTs
8; D) topographic image (a,b and c are respectively the background, the nanotube body and the
decorations).
Chapter 2
39
Figure 10. Phase shift versus bias plot of : background (green), nanotube body (blue) and “decorations”
(red) for f-MWCNTs on silicon oxide. The phase-shift is related to the different regimes: repulsive for
negative phase shift, and attractive for positive phase shifts.
These “decorations” behave differently from the nanotube body in the
phase-shift electrostatic images. To investigate the electrostatic nature of such
areas, we plotted E-EFM phase signals for each of the previously described areas
(silicon oxide background, nanotubes and “decorations”) as a function of different
values of the applied DC voltage. Results are shown in Figure 10. It is evident from
the figure that the “decorations” are in general less attracted by the tip, with
respect to the substrate, at positive bias, and less repulsed at negative bias. This is
compatible with a positive or, at least, less negative than the substrate charge
distribution localized at the “decorations”. More in general, although the presence
of positive charges on MWCNTs, we can assert that the total charge felt by the tip
on the sample (i.e. background + MWCNT+ decorations) is negative, as proved by
the trends of the graphs in Figure 10 (repulsion for negative biases, attraction in
case of positive voltages). This general behavior can be explained, as before, by the
presence of negative terminations distributed along the whole silicon oxide
surface. To better evidence the different contribution of the MWCNTs body and of
the “decorations” to the electrostatic measurements, we have subtracted from the
two curves the background phase signal and the nanotube phase signal (that
contains also the background contribution) respectively. The results are shown in
the inset of graph Figure 10. We can immediately see that the “decorations” behave
differently from the nanotube bodies: the “decorations” trace has a central
symmetry (with respect to the axis origin), compatible with a localized positive
charge, while the nanotube body is symmetric with respect to the y-axis,
compatible with the behavior of a dipole. The comparison of these data to the one
reported for pristine MWCNTs (Figure 8), clearly shows that “swellings” behaves
differently form “decorations” and are more like the CNT body.
Chapter 2
40
A possible explanation for this response is the easy polarizability of carbon
nanotubes due to electron mobility. The asymmetry of the polarization is
reasonably due to the silicon oxide negative charges below the MWCNTs that
reduce the dipole for negative tip voltages.
We can confirm that the nanotubes behave as non-symmetric dipoles, due
to the influence of the underneath silicon oxide surface, while we can discriminate
the presence of positive charges localized at the functionalization sites.
We are currently investigating the electrostatic behavior of SWCNTs with
the same functionalizations, in order to better understand this type of interactions.
2.2 Microwave-Assisted Functionalization of Carbon Nanotubes in
Ionic Liquids2
In this section, an innovative strategy for the functionalization of carbon
nanotubes is presented, designed by the combined use of ILs and microwave
irradiation (Appendix B).
Building on the fullerene experience, a wide variety of reactions have been
carried out for the covalent functionalization of CNTs. At variance with the
fullerene analogues, CNTs require harsh temperature and/or pressure conditions,
with a large use of organic solvents and long reaction times. Therefore, microwave
(MW) assisted and solvent-free protocols have been developed to accelerate
reactions and produce f-CNTs efficiently. A recent achievement has been the
double functionalization of SWCNTs with azomethine ylides and arene radical
additions, both promoted by MW6,7. In this perspective, the optimization of CNTs
functionalization methods should also address the replacement of hazardous VOCs
with alternative reaction media. To this end, ILs have been used successfully for a
2 The work reported in the second part of this chapter is a part of a paper recently accepted for
publication in Chemistry – A European Journal. “Microwave-Assisted Functionalization of Carbon
Nanostructures in Ionic Liquids” by Ivan Guryanov I., Toma F.M., Montellano López A., Carraro M.,
Da Ros T., Angelini G., D’Aurizio E., Fontana A., Maggini M., Prato M., Bonchio M.
Chapter 2
41
variety of organic transformations at room temperature8. Recognized advantages
of ILs are their tuneable composition, polar character, non-volatility, thermal
resistance, complementarity with water or other green media and liquid
electrolyte behavior. In addition, IL phases are instrumental for fast and selective
MW-heating by ionic conduction mechanism, with negligible vapour pressure and
safer set-up conditions9,10. Aggregation phenomena, driven by the low solubility of
CNTs in polar environment, are expected to affect the reactivity behavior in IL
media. Noteworthy, it has been recently demonstrated that SWCNTs, upon
grinding with ILs, form highly viscous gels termed “bucky gels” with extraordinary
mechanical properties, and thermal11, electrochemical12,13 and rheological
properties14 have been investigated by several authors. The gelation refers to
mixtures in which an exfoliation process of the nanotube bundles is induced by the
tendency of ILs to adsorbe around the SWCNT surface. The use of bucky gels as
reaction media has been investigated for the addition of diazonium ions to the
SWCNT surface15. Exfoliation of the SWCNT ropes by the IL environment
accelerates the reaction which occurs in minutes at room temperature.
We investigated the effect of MW irradiation and ILs on the 1,3-dipolar
cycloaddition of azomethine ylides to SWCNTs by screening the reaction protocol
together with the IL medium composition, the concentration factor, the applied
MW power and the simultaneous cooling of the system. This latter turned out to be
crucial for directly dosing high levels of MW power to the reaction without over-
heating16,17. Our results include: (i) the evaluation of the cycloaddition kinetics and
selectivity in both o-DCB and IL containing media; (ii) optimization of the synthetic
protocol towards multi-substitution on the CNTs surface; (iii) extension of the
optimized protocol to fluorous-tagged aldehydes. Under MW-assisted reaction in
IL phases, fluorous-tagged CNTs (FT-CNTs) could be readily isolated. Preliminary
investigations showed a remarkable solubility of FT-CNTs in fluorous phases, with
potential application in catalysis, material science and membrane-based
technology18.
Chapter 2
42
The combined use of MW-activation and ILs provides a new strategy for the
functionalization of purified, non-oxidized SWCNTs by 1,3-dipolar cycloaddition.
We explored the use of the so-called “bucky gels”, as reaction media, when exposed
to reagents and to MW irradiation. We have tested three different ILs, namely 1-
butyl-3-methylimidaziolinium tetrafluoroborate ([bmim]BF4), 1-methyl-3-
octylimidazolinium tetrafluoroborate ([omim]BF4) and 1-hexadecyl-3-
vinylimidazolinium bis(trifluoromethylsulfonyl)imide ([hvim][(CF3SO2)2N])
(Scheme 2) to evaluate the influence of the length of the alkyl chain (4, 8 or 16
carbon atoms, respectively) and of the presence of an additional bond (in the
case of [hvim][(CF3SO2)2N]) on the SWCNT exfoliation ability and cycloaddition
promotion.
Scheme 2. Molecular formulas of [bmim]BF4, [omim]BF4 and [hvim]CF3SO2)2N]
The exfoliation ability has been investigated, in collaboration with
University of Chieti, through rheological measurements at 25 and 60°C in order to
study the changes in stability, consistency and degree of the IL coating on the
SWCNTs by varying the IL nature and the temperature. Indeed, the affinity of ILs
for the SWCNT surface is fundamental for the formation of the gels as pure ILs do
not possess any elastic component. Rheological measurements in the dynamic
mode at a constant shear stress indicate a viscoelastic gel-like behaviour for the
investigated IL-SWCNT mixtures with the measured storage moduli (elastic, G’)
dominating the loss moduli (viscous, G”) by one order of magnitude and exhibiting
little frequency dependence over the range of angular frequencies tested.
Chapter 2
43
This piece of evidence highlights the existence of weak interactions within
the gel11. On the other hand, they point to a large number of weak physical cross-
links among the SWCNT bundles mediated by ILs. As a matter of fact cation-
interactions, and also weaker π-π interactions between alkylimidazolium cations
and the sp2-carbon framework of SWCNTs, may orient and trigger clustering of the
surrounding imidazolium ions and consequently interconnect neighbouring
SWCNTs. In addition, an increase in the temperature (from 25 to 60°C) provokes a
decrease of the viscosity and also a further reduction of the endured shear stress.
As far as the comparison of the investigated gel is concerned, the trend of the
complex viscosity *, of the elastic modulus and of the steady shear viscosity
follow approximately the order [hvim][(CF3SO2)2N] > [omim]BF4 ≥ [bmim]BF4. The
higher the values of G’ and G”, the greater the number of interactions and/or the
stronger the interactions within the gel19. This evidence points to the fact that,
besides the prevailing cation- interactions as already highlighted elsewhere11,
Van der Waals interactions between ILs and SWCNTs surface, due to the presence
of the alkyl chains on the imidazolyl group, do affect the viscosity of the final gels.
Nevertheless, additional - interactions cannot be excluded in the case of
[hvim][(CF3SO2)2N].
Table 1. Rheological characterization of the IL-SWCNT gels.
IL[a] G’/Pa[b] at 25°C
(at 60°C)
G”/Pa[b] at 25°C
(at 60°C)
*cP[b,c] at 25°C
(at 60°C)
y/Pa at
25°C (at
60°C)
cP[d] at 25°C
(at 60°C)
[bmim]BF4 550 (300) 55 (50) 8.9·105 (4.3·105) 12 (5) 5000 (1000)
[omim]BF4 800 (200) 80 (50) 13·105 (3.6·105) 25 (4) 7200 (1800)
[hvim][(CF3SO2)2N] (2500) (400) (43·105) (9) (2400)
[a] Pure [bmim]BF4, [omim]BF4 and [hvim] ][(CF3SO2)2N] have a viscosity of 80 cP at 25°C and 70 cP at 60°C, 240 cP at 25°C and 200
cP at 60°C and 200 cP at 60°C, respectively. [b] Shear stress () 1.0 Pa satisfying the linear viscoelasticity; angular frequency ()
varied from 0.1 to 10 Hz. [c] Calculated at the frequency of 0.1 Hz. [d] Calculated at 10 s-1
.
Chapter 2
44
The gel phase provides the CNTs with a uniform coating of IL film, which
favors exfoliation and assists the covalent modification of the carbon surface
(Scheme 3)20,21. In this scenario, the pseudo-supported IL layer acts as a sort of
extracting medium to: (i) solubilize and transfer the starting reagents, (ii) promote
the in-situ formation of active intermediates, (iii) stabilize uphill pathways and
transition states.
Scheme 3. Azomethine ylide cycloadditions to IL-containing bucky gels under MW irradiation
Furthermore, the IL matrix sets forward a molecular heater array, under
MW irradiation, to foster a flash and highly efficient thermal activation for
endothermal processes. With this aim, we have investigated the 1,3 dipolar
addition of azomethine ylides to SWCNTs in IL-structured gels (Table 2). In the
first instance, the degree of CNT functionalization (i.e. the number of attached
functions per carbon atom of the nanotube) has been established by TGA that
enables us to evaluate the total mass loss as derived from the functional groups
covalently attached to the CNT surface, after subtraction of the pristine SWCNT
contribution. The f-SWCNTs have been further characterized with a variety of
techniques, including Raman, UV-Vis-NIR spectroscopies and TEM. The MW-
assisted protocol has been compared with the conventional synthesis in either
Chapter 2
45
molecular solvent (DMF) or [bmim]BF4 performed for 5 days at 120 °C (entries 9
and 10 in Table 2). The loss of weight were calculated with respect to the starting
material (SWCNTs CNI®).
Table 2. MW-assisted azomethine ylide cycloaddition to SWCNTs in IL.
entry[a] Aldehyde Solvent/T[b] TGA wt loss %[c] N[e]
9[e]
CH3-(CH2)5CHO DMF, 120°C 7 1/164
10[e]
CH3-(CH2)5CHO [bmim]BF4, 120°C 7 1/156
11[f]
CH3-(CH2)5CHO [bmim]BF4, MW 9 1/120
12[f]
CH3-(CH2)5CHO [omim]BF4, MW 6 1/191
13[f]
CH3-(CH2)5CHO [hvmim][(CF3SO2)2N], MW 0 0
14[f],[g]
CH3-(CH2)5CHO [bmim]BF4/o-DCB, MW 16 1/60
15[f],[h]
C8F17-(CH2)2CHO [bmim]BF4, MW 14 1/239
16[f],[i]
C8F17-(CH2)2CHO [bmim]BF4, MW 16 1/211
17[h],[l]
C8F17-(CH2)2CHO [bmim]BF4, MW 27 1/108
[a] In all reactions: SWCNT Carbon Nanotechnologies Incorporated, 5-10 mg, sarcosine 1.0 mmol, aldehyde 2.0 mmol. [b] Solvent
or IL gel and conditions with conventional heating (°C) or MW irradiation under magnetic stirring and simultaneous cooling by
compressed air (MW). [c] % Weight loss determined by thermogravimetric analysis [d] SWCNT functionalization coverage per
carbon atom. [e] Sarcosine 0.2 mmol, aldehyde 0.2 mmol, 5 day reaction. [f] 20 W, Tbulk=120 °C, 1h. [g] [omim]BF4/o-DCB=1/3. [h]
aldehyde 0.2 mmol, sarcosine 0.2 mmol. [i] Aldehyde 0.5 mmol, sarcosine 0.5 mmol. [l] 50 W, Tbulk=140 °C, 1h.
Noteworthy, MW irradiation, for 1h, at 20 W is instrumental for good to
excellent functionalization of SWCNTs, yielding group coverage up to 1/60. This
result represents one step forward in the chemistry of CNTs, as far as product
yields and reaction time are concerned6,7. Moreover, under the conditions adopted,
the bucky gels respond properly to MW absorption, with a well-behaved and
reproducible temperature/pressure raise profile, so that the scaling-up of the
synthesis is feasible with a controlled and safer reaction set-up. It is known that
pristine SWCNTs, under solvent-free MW irradiation, undergo intense heat release
in conjunction with uncontrolled spark emission and ignition phenomena. In a
dense/viscous solvent environment, such highly effective MW absorption is
Chapter 2
46
prevented, but also the flash MW-induced thermal activation is suppressed6,7. In
this light, the bucky gel matrix does represent a powerful alternative, as the IL
medium shields the strong SWCNT response while providing a safer source of
instant heating, by virtue of its fully ionic character.
Data in Table 2 allow to address the impact of the IL nature and of o-DCB as co-
solvent on the reaction efficiency. In particular, an increased performance is
registered in low viscosity bucky gels. Indeed, the functionalization coverage
decreases in the order [bmim]BF4> [omim]BF4> [hvim](CF3SO2)2N (entries 11-13
in Table 2), thus ranking the IL in the reverse order with respect to the viscosity
determination discussed previously. The different ability of the investigated bucky
gels to function as solvent for the dipolar cycloaddition probably stems from the
different mesoscopic structure of the gels formed. Indeed, many gels contain
domains whose microviscosity is close to that of the neat liquid component (the IL
in this case)22, leading to relatively short translational diffusion coefficients despite
a very high macroscopic viscosity. Indeed, in the present cases, the measured G’
and G” moduli clearly highlight increased interactions within the gel on increasing
the length of the alkyl chain, with [hvim][(CF3SO2)2N] being the IL characterized by
the greatest number of interactions and/or the strongest interactions. Therefore,
in the latter case, the mobility of one or more reactants might be particularly
reduced, e.g. by selective entrapment in the gel network, to an extent adversely
affecting the reaction rate.
TEM characterization of entries 9, 12 and 13 in Table 2 (Figure 11) clearly
prove the exfoliation effect of ILs with respect to entry 9, improving dispersibility
independently from the degree of functionalization.
Chapter 2
47
Figure 11. TEM images of f-SWCNTs, obtained as described in Table 2 and dispersed in DMF (1 mg/mL):
A) entry 9; B) entry 12; C) entry 13. TEM bar: 500 nm.
TGA and UltraViolet-Visible-NearInfraRed (UV-Vis-NIR) graphics for entries 10, 12
and 13 are reported in Figure 12.
Figure 12. A) TGA curves of SWCNTs CNI® (red solid line) and f-SWCNTs, obtained as described in Table 2:
entries 10 (black dash dot line), 12 (green dash line), 13 (blue dot line). B) UV-NIR spectra of f-SWCNTs,
obtained as described in Table 2: entries 10 (black dash dot line), 12 (green dash line), 13 (blue dot line).
A
A B
B C
Chapter 2
48
Figure 13. Time evolution of Raman spectra collected for pristine SWCNT and for functionalized samples
under MW-irradaiation: SWCNTs CNI® (red solid line), entry 11 table 2 (black dash dot line), entry 14
table 2 (green dash line). In the inlet, it is visible the increase of the Disorder (D) band: higher is the
functionalization degree, higher is the D band in Raman spectrum.
Figure 14. A) TGA curves of SWCNTs CNI® (red solid line) and of f-SWCNTs, obtained as described in
Table 2: entries 9 (blue dot line), 11 (black dash dot line), 14 (green dash line). B) UV-NIR spectra in DMF
of SWCNTs CNI® (red solid line), entry 9 Table 2 (blue dot line), entry 11 Table 2 (black dash dot line),
entry 14 Table 2 (green dash line).
A B
Chapter 2
49
The addition of o-DCB as a co-solvent brings about a substantial
improvement of the SWCNT functionalization yield (entry 14 in Table 2). This
result is unequivocally confirmed by the time-evolution of the Raman spectra,
collected with excitation at 632.8 nm, of the SWCNTs treated by MW irradiation
under the reaction conditions (Figure 13). The comparison with the Raman
features of the starting material, showing a small disorder mode at 1314 cm-1,
points to a gradual increase of the D-band corresponding to the functionalization
progress. Inspection of the radial breathing mode (RBM) spectral region (<400 cm-
1) provides direct evidence that the applied MW-assisted protocol is not affecting
the overall distribution of the nanotube types. As a further remark, the partial loss
of the characteristic interband transitions between van Hove singularities of
SWCNTs is also consistent with the occurrence of their functionalization (Figure
14).
TEM images indicate that the nanotube bundles are well-dispersible in polar
organic solvents, like DMF (Figure 15).
Figure 15. TEM images of: A) derivative 10, B) derivative 11 and C) derivative 14, obtained as reported in
Table 2. TEM bar: 500 nm. Samples were dispersed in DMF.
The rational explanation for the enhanced efficiency observed with the molecular
co-solvent is likely twofold, and not only related to solubilization/diffusion effects,
but also due to the parallel inhibition of the competitive retro-cycloaddition
process. Indeed, pyrrolidine cycloreversion has been demonstrated also in the case
of covalently modified SWCNTs, under thermal and catalytic conditions4.
A
B
C
Chapter 2
50
The MW-assisted protocol has been applied to the introduction of fluorous-
tagged pyrrolidine moieties onto the SWCNT surface (entries 15-17 in Table 2).
Interestingly a 1/108 functional coverage is obtained by using a lower excess of
the fluorous-tagged aldehyde, at an applied MW power of 50W. The resulting
material features an unprecedented affinity for fluorinated phases. Solutions up to
1 mg in 1 ml of solvent, remain stable for weeks, without showing appreciable
material precipitation (Figure 16). TGA data and TEM images provide converging
evidence of the expected functionalization (Figure 17 and Figure 18).
Figure 16. Pictures of vials containing FT-CNTs in hexafluoro-isopropanol, obtained as described in Table
2: a) entry 15: suspension with 0.5 mg/mL; b) entry 15: suspension with 0.1 mg/mL; c) entry 17: solution
with 0.5 mg/mL.
Figure 17. TEM images of FT-CNTs, obtained as described in Table 2 and dissolved (1 mg/mL) in
hexafluoro isopropanol: A) entry 15; B) entry 16; C) entry 17. TEM bar: 500 nm.
A
A B C
B C
Chapter 2
51
Figure 18. TEM images of FT-CNTs, obtained as described in Table 2 and dissolved (1 mg/mL) in DMF: A)
entry 15; B) entry 16; C) entry 17. TEM bar: 500 nm.
Figure 19. A) TGA curves of SWCNT CNI® (red solid line) and of f-SWCNTs, obtained as described in Table
2: entries 15 (black dash dot line), 16 (green dash line), 17 (blue dot line). B) UV-NIR spectra of FT-CNTs,
obtained as described in Table 6: entries 15 (black dash dot line), 16 (greendash line), 17 (blue dot line).
TGA and UV-NIR spectra of entries 15, 16 and 17 (Table 2) are reported in Figure
19. Thus, the loss of van Hove singularities, especially for entries 8 and 9 is
consistent with the proposed degree of functionalization.
The reported results23 have paved the way for more complex syntheses to
be performed under MW irradiation.
References:
(1) Campidelli, S.; Sooambar, C.; Diz, E.; Ehli, C.; Guldi, D.; Prato, M. Dendrimer-
functionalized single-wall carbon nanotubes: Synthesis, characterization, and
A
B
C
A
B
Chapter 2
52
photoinduced electron transfer. Journal of the American Chemical Society
2006, 128, 12544-12552.
(2) Oh, S.; Kim, Y.; Ye, H.; Crooks, R. M. Synthesis, Characterization, and Surface
Immobilization of Metal Nanoparticles Encapsulated within Bifunctionalized
Dendrimers. Langmuir 2003, 19, 10420-10425.
(3) Sarin, V.; Kent, S.; Tam, J.; Merrifield, R. Quantitative monitoring of solid-
phase peptide synthesis by the ninhydrin reaction. Analytical Biochemistry
1981, 117, 147-157.
(4) Langa, F.; De La Cruz, P. Microwave irradiation: An important tool to
functionalize fullerenes and carbon nanotubes. Combinatorial Chemistry and
High Throughput Screening 2007, 10, 766-782.
(5) Loupy, A. Microwaves in Organic Synthesis; 2° ed.; Wiley-VCH, 2006.
(6) Brunetti, F.; Herrero, M.; Muñoz, J.; Díaz-Ortiz, A.; Alfonsi, J.; Meneghetti, M.;
Prato, M.; Vázquez, E. Microwave-induced multiple functionalization of
carbon nanotubes. Journal of the American Chemical Society 2008, 130, 8094-
8100.
(7) Brunetti, F.; Herrero, M.; Muñoz, J.; Giordani, S.; Díaz-Ortiz, A.; Filippone, S.;
Ruaro, G.; Meneghetti, M.; Prato, M.; Vázquez, E. Reversible microwave-
assisted cycloaddition of aziridines to carbon nanotubes. Journal of the
American Chemical Society 2007, 129, 14580-14581.
(8) Martins, M.; Frizzo, C.; Moreira, D.; Zanatta, N.; Bonacorso, H. Ionic liquids in
heterocyclic synthesis. Chemical Reviews 2008, 108, 2015-2050.
(9) Leadbeater, N.; Torenius, H.; Tye, H. Microwave-promoted organic synthesis
using ionic liquids: A mini review. Combinatorial Chemistry and High
Throughput Screening 2004, 7, 511-528.
(10) Habermann, J.; Ponzi, S.; Ley, S. Organic chemistry in ionic liquids using non-
thermal energy-transfer processes. Mini-Reviews in Organic Chemistry 2005,
2, 125-137.
Chapter 2
53
(11) Fukushima, T.; Kosaka, A.; Ishimura, Y.; Yamamoto, T.; Takigawa, T.; Ishii, N.;
Aida, T. Molecular ordering of organic molten salts triggered by single-walled
carbon nanotubes. Science 2003, 300, 2072-2074.
(12) Fukushima, T.; Asaka, K.; Kosaka, A.; Aida, T. Fully plastic actuator through
layer-by-layer casting with ionic-liquid-based bucky gel. Angewandte Chemie
- International Edition 2005, 44, 2410-2413.
(13) Takeuchi, I.; Asaka, K.; Kiyohara, K.; Sugino, T.; Terasawa, N.; Mukai, K.;
Fukushima, T.; Aida, T. Electromechanical behavior of fully plastic actuators
based on bucky gel containing various internal ionic liquids. Electrochimica
Acta 2009, 54, 1762-1768.
(14) Kim, H.; Choi, J.; Lim, S.; Choi, H.; Kim, H. Preparation and nanoscopic internal
structure of single-walled carbon nanotube-ionic liquid gel. Synthetic Metals
2005, 154, 189-192.
(15) Price, B.; Hudson, J.; Tour, J. Green chemical functionalization of single-walled
carbon nanotubes in ionic liquids. Journal of the American Chemical Society
2005, 127, 14867-14870.
(16) Berardi, S.; Bonchio, M.; Carraro, M.; Conte, V.; Sartorel, A.; Scorrano, G. Fast
catalytic epoxidation with H2O2 and [γ-SiW10O36(PhPO)2]4- in ionic liquids
under microwave irradiation. Journal of Organic Chemistry 2007, 72, 8954-
8957.
(17) Carraro, M.; Sandei, L.; Sartorel, A.; Scorrano, G.; Bonchio, M. Hybrid
polyoxotungstates as second-generation POM-based catalysts for microwave-
assisted H2O2 activation. Organic Letters 2006, 8, 3671-3674.
(18) Tasis, D.; Tagmatarchis, N.; Bianco, A.; Prato, M. Chemistry of carbon
nanotubes. Chemical Reviews 2006, 106, 1105-1136.
(19) Kim, J.; Kim, D.; Kim, S. Effect of modified carbon nanotube on physical
properties of thermotropic liquid crystal polyester nanocomposites.
European Polymer Journal 2009, 45, 316-324.
Chapter 2
54
(20) Tu, W.; Lei, J.; Ju, H. Functionalization of carbon nanotubes with water-
insoluble porphyrin in ionic liquid: Direct electrochemistry and highly
sensitive amperometric biosensing for trichloroacetic acid. Chemistry - A
European Journal 2009, 15, 779-784.
(21) Wu, B.; Hu, D.; Kuang, Y.; Liu, B.; Zhang, X.; Chen, J. Functionalization of
carbon nanotubes by an ionic-liquid polymer: Dispersion of Pt and PtRu
nanoparticles on carbon nanotubes and their electrocatalytic oxidation of
methanol. Angewandte Chemie - International Edition 2009, 48, 4751-4754.
(22) Skrzypczak, A.; Neta, P. Diffusion-controlled electron-transfer reactions in
ionic liquids. Journal of Physical Chemistry A 2003, 107, 7800-7803.
(23) Guryanov I.; Toma F. M.; Montellano López A.; Carraro, M.; Da Ros T.; Angelini
G.; D'Aurizio E.; Fontana A.; Prato M.; Bonchio M. Microwave-Assisted
Functionalization of Carbon Nanostructures in Ionic Liquids. Chemistry - A
European Journal.
Chapter 3
55
Chapter 3:
MWCNT dendron series
for efficient sirna
delivery
Biological studies reported in this chapter were done in collaboration with Dr.
Khuloud T. Al-Jamal and Prof. Kostas Kostarelos from the Nanomedicine
Laboratory, Centre for Drug Delivery Research, The School of Pharmacy, University
of London; and Dr Alberto Bianco from CNRS, Institut de Biologie Moléculaire et
Cellulaire, Immunologie et Chimie Thérapeutiques, Strasbourg, France.
3.1 Drug delivery in cancer nanotechnology: siRNA
After cardiovascular pathologies, cancer is one of the most important diseases
and affects a large portion of the western world. The etiology of this disease,
which is comprehensive of a numbers of different typologies, is related to
genetic instability and/or multiple DNA alterations. One of the main problems is
to selectively distinguish between healthy and tumor cells, and in this respect
compared to current approaches in cancer treatment, nanotechnology's recent
Chapter 3
56
achievements are promising major breakthrough in patient care. As the U.S.
National Cancer Institute states1, nanotechnology offers the unprecedented
opportunity to study and interact with normal and cancer cells in real time, at
molecular and cellular scales, and during the earliest stages of the cancer
process. Nanoscale molecular structures are similar in size to large
biomolecules and, under the right conditions, they can easily enter cells or pass
the blood vessels.
The biomedical nanotechnology approach (namely nanomedicine) can be
really useful in cancer therapy at three different stages of the tumoral process:
early detection, imaging and therapeutic treatment of tumors2. For instance,
delivery system gives us the possibility to target therapeutic or contrast agents
directly to the desired organ or tissue using nanovectors. They tipically consist
of a core decorated on the surface with targeting molecules and/or solubilising
agents, and bearing inside therapeutic agent or imaging payloads. They, in fact,
are engineered to release the loaded units after reaching the target, carring out
their activity only or mainly there, strongly decreasing the side effects. The
recognition units can generally be molecules, which are recognized by receptors
expressed (or in a favorable case over-expressed) on the surface of cells in
tumor tissues. The targeting can take place by an active recognition, due to the
presence of a special unit (active targeting), but can also happen passively, by
natural cellular uptake being, in this case, strictly related to enhanced
permeability and retention effect (EPR). EPR concern tumor angiogenesis,
where blood vessels bear numerous and larger fenestration in cancerous than in
normal tissues, due to the demands of a rapid growth. The endothelium of
cancer blood vessels is, in fact, characterized by the presence of few perycytes
and smooth muscle cells and this causes the shape of fenestrations2. Moreover,
usually in tumor tissue there is a lack of lymphatic vessels and these two
characteristics, with the consequent cell expression of receptors for growth
factors and cell adhesion, allow the use of multiple targeting strategies that
underwent extravasation and protracted lodging3. In this scenario,
Chapter 3
57
reticuloendothelial system (RES), as a part of the immune system and consisting
mostly in macrophages and monocytes localized in lymph nodes, spleen and
Kupffer cells, plays a determinant role because, tuning their size, nanovectors
can target or can escape its cells.
The approval of AbraxaneTM by FDA has been the major success for cancer
nanotechnology in the last years. It consists of paclitaxel coupled to albumin as
delivery agent. This formulation has been successfully adopted in the therapy of
metastatic breast cancer4. Generally, the advantages of using nanovectors are
the possibility to overcome poor solubility of some drugs, to protect them from
metabolic and immune system attacks, to alter the biodistribution by tuning the
nanovector size (exploiting the EPR effect), and to control the release profile by
coupling with enzyme sensitive linkers.
Pharmacokinetics and pharmacodynamics are also important
parameters in studying this new carriers, and in vivo biodistribution studies
represent one of the necessary steps in the development of vectors.
RNA interference mechanism is crucial to study gene functions and to
defeat many diseases as genetic diseases, HIV, influenza or cancer. siRNA consits
of a double stranded RNA that leads to the degradation of the complementary
endogenous mRNA and it is responsible for knocking down specific genes
related to illness5. RNA interference (RNAi) mechanism was discovered by
Jorgensen in 19906 and the related theory was improved by Mello in 19985: they
proved that dsRNA injected in C. Elegans was able to knock down genes. RNAi is
based on DICER, RNAse III able to cut the dsRNA in small nucleotides, i.e. small
interference RNAs (siRNAs). siRNA has a length of about 20-25 ribonucleotides
with two nucleotides overhanging at 3’ end. This peculiarity enables the siRNA
to interact with the RNA-induced silencing complex (RISC) inducing gene
silencing and cleavage of mRNA. The strand incorporated in the RISC is a guide
that activates the enzymatic complex to selectively degrade the complementary
mRNA. The use of siRNA is quite new and it was introduced in 2002 by T.
Tuschl7, and it constitutes an improvement with respect to dsRNA, which
Chapter 3
58
induces an immunogenic response, while the siRNA does not activate this
response.
Even though siRNA could bypass the cellular defense mechnism because
of its length, it needs to be delivered inside cells because it bears many negative
charges and it is not able to cross cell membrane by diffusion. Viral vectors are
widely used but they imply many risks in safety and involve inflammatory and
immunogenic effects. This is why non-viral vectors are preferable and many
cationic derivatives (such as cationic polymers, dendrimers or peptides) are the
most quoted candidates to translocate siRNA across cell membrane. Moreover,
the main part of the existing carriers has been developed to deliver DNA into the
nucleus and they usually undergo to normal endocytosis and subsequent
endolysosomal processing8.
The first attempt of gene silencing using carbon nanotubes was reported
by Dai et al9 in 2005. They f-SWCNTs using non-covalent approach with
phospholipidic-polyethilenglycole chains and attached DNA or siRNA by
disulfur bond showing the translocation of these complexes inside the cells.
Afterwards, Zhang et al10 gave the first proof of silencing the expression of
transcriptase, using a positively charged SWCNT derivative. At the same time,
also our group started working on this subject reporting that plasmid DNA can
be delivered to induce recombinant expression of genes11. More recently, we
demonstrated the ability of MWCNTs 10 to deliver siRNA into human tumor
human xenograft model in vivo12.
3.2 CNTs and Dendrimers as drug carriers
We have previously discussed about citotoxicity and the possible biological
applications of CNTs (Chapter 1) together with their ability to enter cells13 and
their biocompatibility once functionalized14,15. More in detail, carbon nanotubes
own an high aspect ratio, and the large surface area can be exploited to load
different compounds at once. This represents a powerful outcome to escape
Chapter 3
59
multi-drug resistance effects that are often developed by tumor cells16. Few
examples have been reported in our group by Pastorin et al17,18, presenting the
double functionalization of CNTs with Methotrexate and a fluoresceine probe.
So, we decided to couple the advantages of using CNTs with those of using
dendrimers. These are also promising cargo for drug delivery because they bear
a lot of functionalities where drugs or solubilizing agents can be attached.
Nevertheless, the biocompatibility constitutes one of the main concern and
many studies have been done19. Toxicity depends on three different factors: i)
internal core, ii) dendrimer generation, and iii) surface charge. In detail, at high
doses only generations higher than G7 show remarkable toxicity and
immunogenicity20. Noteworthy, PAMAM has been extensively studied for
biological application and two commercial derivatives (Polyfect® and
Superfect®) are available as gene transfecting agent. It seems, indeed, that
amino terminating PAMAM condenses DNA, and protects it when it is
circulating in blood stream21.
Moreover, Fischer et al studied the cytotoxicity of polycation reporting a rank in
toxicity of amine, i.e. primary amine > secondary amine > tertiary amine22.
Herein, we used positively charged MWCNT derivatives with a second
generation PAMAM dendron to complex a siRNA toxic sequence and deliver it to
transfect mammalian cells.
3.3 MWCNTs dendron series for an efficient siRNA delivery1
The alkylated dendron-MWCNT adducts, reported in Chapter 2 (MWCNTs 2,
MWCNTs 5 and MWCNTs 8), were complexed with a non-coding siRNA
sequence and the complexes were successfully used to transfect mammalian
1 Part of the work reported in this chapter has been published in: Journal of American Chemical Society, 2009, 131 (28), pp 9843-9848. “Synthesis and characterization of a carbon nanotube-dendron series for efficient siRNA delivery” by Herrero M.A., Toma F.M., Al-Jamal K.T., Kostarelos K., Bianco A., Da Ros T., Bano F., Casalis L., Scoles G., Prato M. One more manuscript is currently in preparation.
Chapter 3
60
cells. We have discovered that the cytoplasmatic delivery of siRNA remarkably
increases with the dendron generation (from G0 to G2) if compared to the
ammonium f-MWCNT precursors (MWCNTs 1). In this study we used a
different batch of CNTs, with a different degree of functionalization if compared
to that one reported in Chaper 2. The whole dendron series were prepared
starting from the same batch of CNT (MWCNTs 1), to achieve comparable
results in degree of functionalization. Figure 1 and Table 1 reported respectively
the TGA analyses and the degree of functionalization of the dendron, calculated
by TGA and Kaiser Test.
Figure 1. TGA curves of alkylated (G0) dendron MWCNTs 2 (red solid line), alkylated (G1) dendron-
MWCNT 5 (blue dot line), alkylated (G2) dendron-MWCNT 8 (green dash line). All the experiments were
performed under N2 atmosphere.
Chapter 3
61
Table 1. The weight percentage was calculated at 450°C.
%Weight Funct/Ca mmol/g
b mmol/g
c
Alkylated (G0)
dendron MWCNT 2 98.6 1915 0.0429 0.072
Alkylated (G1)
dendron MWCNT 5 96.2 1476 0.0543 0.111
Alkylated (G2)
dendron MWCNT 8 94.4 2049 0.0384 0.212
(a) Number of functional groups per each carbon atom and (b) mmol/g calculatied by means of TGA analysis; (c) Kaiser test’s
results.
The data reported in Table 1 show a good correlation between the generations of
the alkylated dendrons.
3.3.1 (G2)dendron-MWCNT:siRNA complexation
As previously shown23, cationic carbon nanotubes are able to efficiently
condense nucleic acids and to deliver them into mammalian cells. Complexation
between the polycationic alkylated (G2) dendron-MWCNT 8 and siRNA was
observed by agarose gel electrophoresis (Figure 2A).
Chapter 3
62
Figure 2. (A) Electrophoretic motility of MWCNT 8 (reconstituted at 1.0 mg/ml in 10% dextrose)
complexed with siRNA. Various mass ratios of the MWCNT 8 were complexed to a fixed 0.5 µg siRNA; (B)
TEM images of MWCNT 8 alone or (C) complexed to siRNA at 1:16 (MWCNT:siRNA) mass ratio . The
scale bar is 500 nm.
The amount of free siRNA migrating into the gel was reduced by increasing the
alkylated (G2) dendron-MWCNT:siRNA mass ratio to almost no free siRNA for 64:1
mass ratio, as evidenced by the reduction of the ethidium bromide fluorescence
intensity in the last two lanes. These results are consistent with previous studies
performed in our group12. The interactions between the siRNA duplex and the
alkylated (G2) dendron-MWCNT 8 were also shown using TEM. In this case mass
ratios lower than 64:1 were used. Figures 2B and 2C show the association of siRNA
around the MWCNTs, which are highly dispersed. The complexation is evident by
the formation of highly electron dense areas (darker in TEM) around the alkylated
(G2) dendron-MWCNT 8 and clustering of the nanotubes presumably due to
bridging by the negative siRNA.
Chapter 3
63
3.3.2 Effect of the different generation alkylated dendron-
MWCNT:siRNA cellular uptake
To establish if the complexation of different generation alkylated dendron-
MWCNTs with siRNA had any effect on the cellular uptake of siRNA, human
cervical carcinoma (HeLa) cells were incubated with the dendron-MWCNT
complexed with a non-coding siRNA sequence. In order to observe whether the
siRNA was internalized in cells, it was fluorescently labelled at the 3´-end with
ATTO 655 emitting at 690 nm. The complexes were prepared using a dendron-
MWCNT:siRNA mass ratio 16:1 as previously optimized12, and cellular uptake was
observed using confocal microscopy under identical optical conditions. Very
interestingly, all cells remained intact with no signs of toxicity after incubation
with the complexes (see Appendix B for details).
Chapter 3
64
Figure 3. Confocal microscopy and differential interference contrast (DIC) images of HeLa cells obtained
after 24 hr incubation with (a) ATTO 655-labeled siRNA alone, ATTO 655-labeled siRNA complexed to (b)
MWCNT 1, (c) alkylated (G0) dendron-MWCNT 2, (d) alkylated (G1) dendron-MWCNT 5 and (e) alkylated
(G2) dendron-MWCNT 8 at 16:1 mass ratio and 80 nM siRNA concentration (equivalent to 16 μg/ml
dendron-MWCNT). Cells were incubated with the complexes for 4 hrs in serum-free media after which
serum-containing media was added to make the final concentration of serum to 10%. Nuclei were
Chapter 3
65
counterstained with propidium iodide. Images show an increase in the intracellular uptake of siRNA as
shown by the increase in the blue fluorescence (siRNA) in the following order e > d> c >b. siRNA alone
gave almost no detectable fluorescence in the cells (a).
Figure 3 shows that dendron-MWCNTs with increased degree of branching
and alkylated dendron generations exhibited more effective intracellular siRNA
delivery than MWCNT 1, while almost no uptake of siRNA alone was observed
(Figure 3a). The fluorescent signal from ATTO 655-labelled siRNA was found
diffused throughout the cytoplasm of cells treated with the dendron-
MWCNT:siRNA complexes, without any indication of localization in intracellular
vesicles (usually evidenced by highly fluorescent pockets within the cell). The
morphology of the cells did not seem to be affected by the dendron-MWCNT,
indicating that the uptake of siRNA was not a result of cellular damage, but
presumably due to the translocation of the dendron-MWCNT through the plasma
membrane. Confocal microscopy studies at early time points (4 hrs) after
incubation of the f-MWCNT:siRNA complexes indicated fluorescence signals of
lower intensity compared to the 24hr data shown here. At both time points the
fluorescent signal from the siRNA-ATTO655 was observed throughout the whole
cell volume without localization in the nucleus or other cellular compartment.
However, the cellular internalization of the siRNA was always nanotube-
dependent. In order to have a better understanding of the mechanism by which
dendron and ammonium f-MWCNT:siRNA complexes enter into cells, systematic
cell trafficking studies of these complexes are warranted and are currently under
investigation in our laboratories. This data show that higher dendron generations
grown on the MWCNT surface could lead to more efficient cytoplasmic delivery of
siRNA, indicating that such constructs can potentially reach higher highly effective
levels of gene silencing. Therefore, using a siRNA sequence that can be cytotoxic to
cells only when delivered efficiently, we have shown that internalization of the
alkyalted (G2) dendron-MWCNT:siTOX complexes could lead to biologically active
delivery of the siRNA, and gene silencing evidenced by extensive apoptosis and cell
death.
Chapter 3
66
Thus, we chose MWCNTs 8 for the further studies.
3.3.3 Alkylated (G2) dendron-MWCNT:siRNA delivery and silencing
with a cytotoxic sequence
The cytotoxicity of alkylated (G2) dendron-MWCNT 8 alone was first
evaluated using human lung epithelial cell cultures (A549) incubated with
different doses (2.5-10 µg/ml) for 24 h. The cytotoxicity was evaluated after
incubation (24h) and at 72 h by flow-cytometry using the Annexin V/Propidium
iodide (PI) assay (Figure 4).
Figure 4. Cytotoxicity assays performed on A549 monolayers incubated with media containing 0, 2.5, 5
and 10 µg/ml of dendron-MWCNT 8. Cells were washed and re-incubated with complete media and
assayed for death after 24 (black bars) and 72 h (magenta bars) by Annexin V/propidium iodide assay
using flow cytometry.
No sign of toxicity was observed up to 10 µg/ml after 24 h or 72 h. Similar results
were also obtained by two independent toxicity assays (LDH and MTT; Figure 5)
altogether indicating that the dendron-MWCNT is innocuous at concentrations up
to 40 µg/ml to A549 cells.
Chapter 3
67
Figure 5. (A) LDH test performed on A549 monolayers incubated with media containing 0, 2.5, 5 and 10
µg/ml of dendron-MWCNT 8. (B) MTT test performed on A549 monolayers incubated with media
containing 0, 2.5, 5, 10, 20 and 40 µg/ml of dendron-MWCNT 8.
Annexin V/PI test highlight the presence of apoptotic cells: viable cells are
both Annexin V and PI negative, whereas cells in early apoptosis are only Annexin
V permeable, and apoptotic cells are permeable to both Annexin V and PI. MTT (a
B
Chapter 3
68
tretrazole derivative) test is an enzymatic test to evaluate the activity of
mitochondria. Viable cells are able to convert MTT to a purple colored molecule.
Finally, LDH (lactate dehydrogenase) assay measures the quantity of NADH
oxidized as a function of cell membrane permeability.
3.2.4 Alkylated (G2) dendron-MWCNT:siRNA delivery and silencing
with the use of a cytotoxic sequence
To assess if the dendron-MWCNT:siRNA complexes can be considered a real
alternative to non-viral vector systems for the efficient delivery of siRNA, a
commercial apoptosis-inducing siRNA sequence (siTOX®) was selected. Effective
delivery of siTOX in cells leads to a cytotoxic response and cell death24. HeLa cells
were incubated with free siTOX (80 nM) or complexed with the alkyalted (G2)
dendron-MWCNT 8 at 1:16 mass ratio (equivalent to 16 µg/ml dendron-MWCNT
8) for 24 h and transfection was assessed after 48 h using the terminal
deoxynucleotidyltransferase–mediated nick end labeling (TUNEL) reaction.
Apoptotic cells were green stained because of incorporation of fluorescein-12-
dUTP at 3'-OH using the Terminal Deoxynucleotidyl Transferase onto the ends of
nicked DNA, while propidium iodide counterstained the nuclei in red (Figure 6).
Only few apoptotic cells can be seen in cultures treated with the dendron-MWCNT
8 alone (Figure 6B), in agreement with the cytotoxicity results obtained from
Annexin V/PI assay or with siRNA alone (Figure 6C). The majority of cells treated
with dendron-MWCNT:siTOX were stained in green with a concomitant reduction
in the total number of nuclei, indicating extensive cell death (Figure 6D). Fewer
cells were seen in the alkylated (G2) dendron-MWCNT:siTOX treated panel because
cell death leads to detachment of cells before or during TUNEL assay processing.
Overall, these data indicate that dendron-MWCNT 8 was able to condense,
intracellularly translocate and successfully deliver biologically active siRNA
sequences.
Chapter 3
69
Figure 6. Confocal microscopy images of HeLa cells obtained 48 h after transfection with alkyalted (G2)
dendron-MWCNT 8:siTOX (16:1 mass ratio) using the TUNEL assay. Untreated cells (A) or cells treated
with dendron-MWCNT 8 (B) or siTOX alone (80 nM) (C) exhibited no sign of apoptosis, while cells
treated with siTOX (D) complexed with the alkyalted (G2) dendron-MWCNT 8 show evident signs of
apoptosis (green) and cell death (fewer cells).
3.2.5 In vivo biodistributional study of dendron-MWCNT derivative
In our group, the evaluation of biodistribution of MWCNTs 1 linked to
[111In]DTPA (diethylentriaminopentaacetic acid) was already reported and
urinary excretion was proved25. It is well known that the main excretion way for
drugs is the renal one, in fact an average of 650 mL/min of plasma are filtrated
by renal glomeruli. In this process the molecular weight is predominant, and
drugs up to 60kDa can pass through renal capillaries. Substances, which do not
undergo renal excretion because of their weight, are eliminated by the liver.
Beside molecular weight, other factors govern the elimination process, like pH
or liposolubility of xenobiotics. Nevertheless, the nanoparticles elimination
process can be more complicated and Choi et al studied the renal clearance of
quantum dots, discovering that it is related to the hydrodynamic diameter as
well as to molecular weight26. Preclinical studies are essential for clinical
BA
DC
BA
DC
Chapter 3
70
applications and we reported the in vivo biodistribution of DTPA-MWCNT 7
(Scheme 1) complexed with 111In. In detail, diethylene triamine pentaacetic
anhydride was added to a Dimethylsulfoxide (DMSO)/DMF solution of MWCNT
7 (Kaiser Test of 1097 mol/g, Chapter 2). Thus, the product was complexed
with 111InCl3, a -emitting radionuclide, to obtain [111In]DTPA-MWCNTs 7.
Scheme 1. DTPA-MWCNT 7 synthesis.
TEM image of DTPA-MWCNT 7 is reported in Figure 7, showing that no significant
change occurred after the reaction.
Chapter 3
71
Figure 7. TEM image of DTPA-MWCNTs 7 (Scale bar 100 nm).
50 g of [111In]DTPA-MWCNT 7 were administrated by tail vein injection to
Balb/C mice and after 24h the Single Photon Emission Computed Tomography
(SPECT) was performed to visualize the systemic distribution of MWCNT
derivative. SPECT allows few days biodistribution imaging and 3D projections can
be software reconstructed by different sagittal section images (Figure 8a).
Chapter 3
72
Figure 8. In vivo biodistribution of [111In]DTPA-dendron-MWCNT in Balb/C mice after single dose
administration via tail vein shown after 24hrs by (a) SPECT/CT , or (b) gamma counting expressed as
percentage injected dose (%ID) per gram tissue 24hrs. Data were expressed as means ± SD (n=3-4).
In Figure 8b, biodistribution analysis is reported, highlighting that the majority
of the MWCNT 7 are concentrated in liver and spleen. Spleen is responsible for
mechanical filtration of blood and this is probably the reason why MWCNTs are
accumulated in it, but to clarify this point further studies are necessary, and we are
now functionalizing MWCNTs 8 with fluorescent probe to study more in detail and
to achieve a detailed understanding of the fate of our derivatives. Finally, the
Chapter 3
73
balance between the excretion and retention of CNTs should be thoroughly
studied, particularly following multiple administrations, in order to better
understand the accumulation of f-CNTs inside the body with respect to their rate of
excretion.
References:
(1) Exploring Nanotechnology in Cancer.
(2) Cuenca, A.; Jiang, H.; Hochwald, S.; Delano, M.; Cance, W.; Grobmyer, S.
Emerging implications of nanotechnology on cancer diagnostics and
therapeutics. Cancer 2006, 107, 459-466.
(3) Ferrari, M. Cancer nanotechnology: opportunities and challenges. Nat Rev
Cancer 2005, 5, 161-171.
(4) Nie, S.; Xing, Y.; Kim, G.; Simons, J. Nanotechnology applications in cancer;
2007; Vol. 9, pagg. 257-288.
(5) Fire, A.; Xu, S.; Montgomery, M.; Kostas, S.; Driver, S.; Mello, C. Potent and
specific genetic interference by double-stranded RNA in caenorhabditis
elegans. Nature 1998, 391, 806-811.
(6) Napoli, C.; Lemieux, C.; Jorgensen, R. Introduction of a chimeric chalcone
synthase gene into petunia results in reversible co-suppression of
homologous genes in trans. Plant Cell 1990, 2, 279-289.
(7) Martinez, J.; Patkaniowska, A.; Urlaub, H.; Lührmann, R.; Tuschl, T. Single-
stranded antisense siRNAs guide target RNA cleavage in RNAi. Cell 2002, 110,
563-574.
(8) Gao, K.; Huang, L. Nonviral methods for siRNA delivery. Mol. Pharm 2009, 6,
651-658.
(9) Kam, N.; Liu, Z.; Dai, H. Functionalization of carbon nanotubes via cleavable
disulfide bonds for efficient intracellular delivery of siRNA and potent gene
silencing. Journal of the American Chemical Society 2005, 127, 12492-12493.
Chapter 3
74
(10) Zhang, Z.; Yang, X.; Zhang, Y.; Zeng, B.; Wang, S.; Zhu, T.; Roden, R. B. S.; Chen,
Y.; Yang, R. Delivery of telomerase reverse transcriptase small interfering
RNA in complex with positively charged single-walled carbon nanotubes
suppresses tumor growth. Clin. Cancer Res 2006, 12, 4933-4939.
(11) Singh, R.; Pantarotto, D.; McCarthy, D.; Chaloin, O.; Hoebeke, J.; Partidos, C.;
Briand, J.; Prato, M.; Bianco, A.; Kostarelos, K. Binding and condensation of
plasmid DNA onto functionalized carbon nanotubes: Toward the construction
of nanotube-based gene delivery vectors. Journal of the American Chemical
Society 2005, 127, 4388-4396.
(12) Podesta, J. E.; Al-Jamal, K. T.; Herrero, M. A.; Tian, B.; Ali-Boucetta, H.; Hegde,
V.; Bianco, A.; Prato, M.; Kostarelos, K. Antitumor activity and prolonged
survival by carbon-nanotube-mediated therapeutic siRNA silencing in a
human lung xenograft model. Small 2009, 5, 1176-1185.
(13) Pantarotto, D.; Singh, R.; McCarthy, D.; Erhardt, M.; Briand, J.; Prato, M.;
Kostarelos, K.; Bianco, A. Functionalized carbon nanotubes for plasmid DNA
gene delivery. Angewandte Chemie - International Edition 2004, 43, 5242-
5246.
(14) Smart, S.; Cassady, A.; Lu, G.; Martin, D. The biocompatibility of carbon
nanotubes. Carbon 2006, 44, 1034-1047.
(15) Dumortier, H.; Lacotte, S.; Pastorin, G.; Marega, R.; Wu, W.; Bonifazi, D.;
Briand, J.; Prato, M.; Muller, S.; Bianco, A. Functionalized carbon nanotubes
are non-cytotoxic and preserve the functionality of primary immune cells.
Nano Letters 2006, 6, 1522-1528.
(16) Singh, R.; Lillard Jr., J. Nanoparticle-based targeted drug delivery.
Experimental and Molecular Pathology 2009, 86, 215-223.
(17) Pastorin, G.; Wu, W.; Wieckowski, S.; Briand, J.; Kostarelos, K.; Prato, M.;
Bianco, A. Double functionalisation of carbon nanotubes for multimodal drug
delivery. Chemical Communications 2006, 1182-1184.
(18) Prato, M.; Kostarelos, K.; Bianco, A. Functionalized carbon nanotubes in drug
design and discovery. Accounts of Chemical Research 2008, 41, 60-68.
Chapter 3
75
(19) Svenson, S. Dendrimers as versatile platform in drug delivery applications.
European Journal of Pharmaceutics and Biopharmaceutics 2009, 71, 445-462.
(20) Roberts, J. C.; Bhalgat, M. K.; Zera, R. T. Preliminary biological evaluation of
polyamidoamine (PAMAM) dendrimers. Journal of Biomedical Materials
Research 1996, 30, 53-65.
(21) Yellepeddi, V.; Kumar, A.; Palakurthi, S. Surface modified poly(amido) amine
dendrimers as diverse nanomolecules for biomedical applications. Expert
Opinion on Drug Delivery 2009, 6, 835-850.
(22) Fischer, D.; Li, Y.; Ahlemeyer, B.; Krieglstein, J.; Kissel, T. In vitro cytotoxicity
testing of polycations: Influence of polymer structure on cell viability and
hemolysis. Biomaterials 2003, 24, 1121-1131.
(23) Lu, Q.; Moore, J.; Huang, G.; Mount, A.; Rao, A.; Larcom, L.; Ke, P. RNA polymer
translocation with single-walled carbon nanotubes. Nano Letters 2004, 4,
2473-2477.
(24) Oh, J.; Razfar, A.; Delgado, I.; Reed, R.; Malkina, A.; Boctor, B.; Slamon, D.
3p21.3 tumor suppressor gene H37/Luca15/RBM5 inhibits growth of human
lung cancer cells through cell cycle arrest and apoptosis. Cancer Research
2006, 66, 3419-3427.
(25) Lacerda, L.; Soundararajan, A.; Singh, R.; Pastorin, G.; Al-Jamal, K.; Turton, J.;
Frederik, P.; Herrero, M.; Li, S.; Bao, A.; Emfietzoglou, D.; Mather, S.; Phillips,
W.; Prato, M.; Bianco, A.; Goins, B.; Kostarelos, K. Cover Picture: Dynamic
Imaging of Functionalized Multi-Walled Carbon Nanotube Systemic
Circulation and Urinary Excretion (Adv. Mater. 2/2008). Advanced Materials
2008, 20, NA.
(26) Soo Choi, H.; Liu, W.; Misra, P.; Tanaka, E.; Zimmer, J. P.; Itty Ipe, B.; Bawendi,
M. G.; Frangioni, J. V. Renal clearance of quantum dots. Nat Biotech 2007, 25,
1165-1170.
Chapter 3
76
Chapter 4
77
Chapter 4:
CNT Substrates for
neuronal cultures:
A morphological study
Electrophysiological studies reported in this chapter were done in collaboration
with Miss Giada Cellot and Prof. Laura Ballerini neurophysiologists from the Life
Science Department, B.R.A.I.N., University of Trieste.
4.1 Introduction: a new standard for substrate preparation1
As it has been already mentioned, CNTs have extraordinary conductive
properties (Chapter 1) and their potential role in repairing injured nerves in the
brain and spinal cord has recently been shown by Ballerini, Prato and
collaborators1 and has recently captured wide interest, as shown by its citation
record of approximately 100 citations in 4 years. This research field has thrieved
because of the biocompatibility of functionalized CNTs and especially because their
highly electrically connected structure when used as surface supported substrates.
CNTs coated glasses, indeed, are well established substrates for cell cultures as
they are known to promote cell attachment, differentiation, growth and long term
1 The data reported in this chapter have not been pusblished yet.
Chapter 4
78
survival comparably with those obtained using other growth enhancing factors
(GEF) to coat glass (like poly-ornithine, poly-lysine or matrigel). Moreover a
promising application in this field is related to the chemical functionalization of
CNTs: the common factors used to pre-treat glass substrates are in fact positively
charged, and it has been shown that neuronal growth and development are
enhanced on this kind of scaffold2.
In this thesis we continue the effort started in our group by Lovat et al in
20051 and later by Cellot et al3, testing the impact of CNT substrates on neuronal
network activity by means of dual electrophysiological recordings. We performed
both electrophysiological measurements, and morphological studies of
hippocampal neuronal cultures grown over carbon nanotube substrates to
investigate the interactions between CNTs and neurons. In Lovat et al1 the
occurrence frequency of spontaneous postsynaptic currents (PSCs) was
monitored, as a parameter to assess the network efficacy and the functional
synapses formation of neuron growth on CNT substrates. PSCs are induced in
postsynaptic neurons by neurotransmitters released by presynaptic neuron
immediately after the action potential. Neurons grown on CNTs displayed on
average a six-fold increase in the frequency of PSCs with respect to control (bare
glass substrate). The authors1 proved also that this boosting of network activity
was not related to a higher number of surviving neurons and in general neurons
deposited on CNTs did not exhibit any difference in “basal electrophysiological
properties” with respect to control.
The procedure to prepare CNTs-coated substrates follows mainly the
previous work of Lovat et al1. In addition, we decided to standardize it and to
produce a common protocol that could allow us to vary other parameters, such as
CNT density, length or type of functionalization and making correlations with the
biological response. 1,3 dipolar cycloaddition (Chapter 2) is needed to increase the
solubility of CNTs, facilitating dissolution in DMF and improving the homogeneity
upon the glass substrates. Attempts to prepare substrates with as-produced,
impure, MWCNTs were already reported1, without any result: metal nanoparticles
Chapter 4
79
are present in this case as a residue of CNTs synthesis, and represent a risk for
neuronal survival
MWCNTs 9, reported in scheme 1, were synthesized (as already reported in
Chapter 2, Table 2) and were characterized by means of TGA under air to assess
the metal content (Figure 2).
Scheme 1. 1,3-dipolar cycloaddition, thermal condition on MWCNTs
The metal content was found to be about 3.9 % and it did not perturb cell viability.
Figure 1. TGA under air atmosphere of Pristine MWCNTs (red solid line) and MWCNTs 9 (black
dash dot line)
These nanotubes were characterized also with TEM and their average length was
calculated (Figure 2). The dispersion ranges (graph in Figure 2) from less than 200
nm to 1250 nm (n=91), the calculated average value being 480 nm (± 250 nm).
Chapter 4
80
Figure 2. TEM image of MWCNTs 9 (scale bar 1 micron). Distribution of MWCNTs 9 diameters.
Our first attempt was to produce a protocol to improve reproducibility and,
above all, to avoid possible cause of neuronal toxicity. After other tests, 5 mL of a
water solution (0.04 mg/mL) of MWCNTs 9 were evaporated over 3 glass
substrates that were deposited in a crystallizer (crystallizer surface area = 1809
mm2). The solution was slowly dried out leaving an average density of CNTs per
glass substrate of 1.1x10-4 mg/mm2, calculated as quantity of MWCNTs in the 5 mL
of solution (0.2 mg) divided by the surface area of the crystallizer. The substrates
were then placed in oven at 350 °C under nitrogen atmosphere in order to remove
the organic functionalization. The main drawback of this method is that the
substrate becomes inhomogeneous in coverage, and there are too many “areas”
that do not allow patch clamping measurements. Neuronal cultures were,
Chapter 4
81
nevertheless, produced from dissociated hippocampal neurons obtained from
brains of newborn rats, and leaving the cells growing on MWCNT modified
substrates for 8 days, then glasses were fixed with glutaraldehyde and coated with
about 10 nm of Ni, for SEM imaging (Appendix A). Although neurons showed high
affinity for the substrates, the biological response was smaller than previously
reported1,3 and this was enough to discard this substrate preparation protocol.
To improve the substrate preparation, we optimized the concentration
conditions, choosing to deposit 2 mL by depositing 10 times 0.2 mL of a solution of
MWCNTs 9, at 0.01 mg/mL. The density of the MWCNT film over the glass (glass
surface area = 288 mm2 ) was about 7x10-5 mg/mm2, calculated as an average of
the quantity of MWCNTs divided by the surface area of the glass. Then, we treated
the substrates for 20 minutes in oven under N2 atmosphere (Scheme 2), to remove
the functionalization.
Scheme 2. Schematic representation of MWCNTs depostition over glass substrates and subsequently
defunctionalization.
SEM characterization of these substrates was then performed, as shown in Figure 4
A and B, and their enlargements in Figure 4 C and D.
A
B
C
D
Chapter 4
82
A
B
Chapter 4
83
Figure 4. SEM images of DMF solution MWCNTs 9 coated substrates (scale bar A 200 m; B 20 m; C, D
2 m).
D
C
Chapter 4
84
From Figure 4, it is visible that these substrates are homogenous in the MWCNT
content. On these substrates, the electrophysiological results obtained by Cellot et
al3 were reproduced.
By means of single-cell electrophysiology, Cellot et al3 demonstrated that
direct interaction of CNTs with the neuronal membrane affects each cell activity.
They discovered that carbon nanotubes increase the probability of generating a
depolarization in membrane potential after a series of action potentials induced by
injections of current pulses into the soma of the cell. Such a phenomenon, which is
calcium mediated, is a mark of an increased ability of neurons in integrating back-
propagating action potentials in the distal compartments of dendrites and, in
simpler terms, makes the neurons more excitable. A schematic representation of a
neuron is presented in Figure 5.
Chapter 4
85
Figure 5. Schematic representation of central nervous system neurons4.
The mechanism underlying this phenomenon has not been totally clarified
yet, but the authors proposed the hypothesis that nanotubes, thanks to both the
characteristic nanostructure and the good conductive properties, provide an
electrical short-cut between adjacent regions of soma and dendrites, that favors
the back-propagation of action potentials. Such an idea was also supported by TEM
images, showing the existence of several points of tight contact between carbon
nanotubes and neuronal membranes (Figure 6).
Figure 6. The interaction between MWCNTs and neurons: note how nanotubes are “pinching” neuronal
membranes3. In the figure Neuronal membrane and the CNT are indicated.
4.2 Morphological study of neurons deposited on CNTs
On this basis, we proceeded in studying the morphological interaction
between neurons and carbon nanotubes. As introduced in Appendix A, SEM is
one of the best techniques that allows this kind of investigation.
In the next micrographs, it is clear that neuronal neurites interact with
CNTs. We are not able to distinguish whether nanotubes penetrate the cell
membrane or not, but we can correlate the “functional bridge” (highlighted in
B
CNT
Neuronal membrane
Chapter 4
86
Figure 7 E, F, G and H) with the TEM images reported by Cellot et al where it is
possible to distinguish few nanotubes in contact with the cell membrane.
A
A
B
Chapter 4
87
C
C
D
Chapter 4
88
E
F
E
F
Chapter 4
89
Figure 7. SEM images of hippocampal neurons grown on DMF solution of MWCNT coated substrates: A,
B, C, D and E, F, G, and H are sequential magnifications referred to two different glass substrates (scale
bar A = 2 m; B, F = 1 m; C, D, G, H = 200 nm; E = 10 m) The red square is referred to the areas that
are subsequently enlarged, the arrows indicate the CNT when it is pinching the dendrites.
F
E
G
H
Chapter 4
90
The images shown in Figure 7 corroborate the hypothesis of the existence
of an electrical shortcut between adjacent compartments of soma and dendrites
mediated by the carbon nanotube carpet. In figure 7 the arrow indicates the
possible zones in which one of these short-cuts may occur. This result could be
confirmed using “non-conductive” carbon nanotubes. Actually, the
functionalization breaks the carbon nanotubes electronic structure reducing the
conducting properties. To this aim, our idea is to produce MWCNTs 8 (Chapter 2)
coated substrates in order to study if the depolarization in neuronal membrane
potential due to the integration of back-propagating action potentials is still
present.
In addition, in this way we could use positively charged pre-treated
substrates that, as mentioned, should also promote neuronal growth and
differentiation.
4.3 Pairs of neurons electrophysiology
We pursued also in the electrophysiological side performing pairs of neuron
recordings in order to characterize neuronal activity in presence of carbon
nanotube substrates.
Briefly, the patch clamp technique was introduced in 1976 by Sackman and
Neher: with this technique, currents of pA and potential of mV are measured. A
glass micropipette (with a tip of 1-2 m) filled with ions solution represents the
recording electrode and it is sealed with the neuronal cell membrane.
Two different modes of recording can be used: current clamp configuration
monitors variations in the membrane potential with the possibility of injecting
currents to depolarize (positive current) o hyperpolarize (negative current) the
cell; in this case spontaneous activity of neurons can be measured as a function of
the number of action potentials over the time.
In voltage clamp configuration, it is possible to record currents passing
through channels present in cellular membrane at determined voltage.
Chapter 4
91
In this case, neuronal spontaneous activity is measured in terms of the
number of post-synaptic currents over the time. Such postsynaptic currents are
generated by the binding of neurotransmitters to their receptors, which are ligand-
gated ion channels.
In dual electrophysiological recordings, pairs of cells, far away 150-200 µm
one from the other, were chosen and one of the two cells, the presynaptic neuron,
was stimulated in current clamp configuration while simultaneously the activity of
the second neuron, the postsynaptic one, was monitored in voltage clamp
configuration, to check the presence or the absence of a response due to the
presynaptic stimulus (Figure 8).
All these recordings used an inverted microscope and this is why we need
transparent substrates and transparent carbon nanotubes film to coat the glass
surface.
Figure 8. Dual electrophysiological recordings. Left: Image of a pair of neurons with glass pipettes for
patch clamp recording. Right: Scheme of the experimental setting.
In Figure 9 (left), the example of a pair of neurons that are not monosynaptically
coupled is shown. In this case, there is no response in postsynaptic trace caused by
the presynaptic stimulation. Contrarily, in Figure 9 (right) the recordings relative
to two neurons forming a synapse are reported: for each presynaptic action
potential, it is present a deflection in the postsynaptic trace that represents an
Chapter 4
92
inward current due to the activation of postsynaptic receptors via
neurotrasmitters released at the presynaptic terminal
Figure 9. Examples of traces for pairs of neurons in control condition and in presence of CNT.
The probability of finding coupled pairs in control conditions and in
presence of CNTs was analyzed. The experiments showed that in control (neurons
grown on bare glass) the percentage of monosynaptic coupling is 39% (n=149),
while when neurons are grown on CNT layers the probability is increased to 58%
(n=122).
Chapter 4
93
In other terms, these data indicate CNT substrates favor the formation of synaptic
contacts between neurons. Such result agrees with one previously reported by
Lovat et al, 2005, where an increased number of synaptic connections promotes
the boosting of spontaneous activity.
To support our previous statements, we present herein also the result
obtained with water dissolved MWCNTs in which is demonstrated that the
probability of finding coupled pairs is higher with respect to control, but lower
Chapter 4
94
with respect to DMF dissolved MWCNTs (Coupled pairs in control: 36% n=28; in
presence of water dissolved MWCNTs: 53% n=30).
This allowed us to conclude that also CNT substrate density plays a role in
neuronal network activity.
Dual electrophysiological recordings demonstrate that the connectivity of
neurons is improved in the presence of carbon nanotubes supporting the results
obtained by Lovat et al1 We are still investigating which is the real role played by
CNTs in the modification of neuronal activity and thanks to this pairs recording we
will highlight whether morphological or electrical properties of these versatile
nanostructures are determining the boosted activity of neural network.
References
(1) Lovat, V.; Pantarotto, D.; Lagostena, L.; Cacciari, B.; Grandolfo, M.; Righi, M.;
Spalluto, G.; Prato, M.; Ballerini, L. Carbon Nanotube Substrates Boost
Neuronal Electrical Signaling. Nano Letters 2005, 5, 1107-1110.
(2) Sucapane, A.; Cellot, G.; Prato, M.; Giugliano, M.; Parpura, V.; Ballerini, L.
Interactions Between Cultured Neurons and Carbon Nanotubes: A
Nanoneuroscience Vignette. Journal of Nanoneuroscience 2009, 1, 10-16.
(3) Cellot, G.; Cilia, E.; Cipollone, S.; Rancic, V.; Sucapane, A.; Giordani, S.;
Gambazzi, L.; Markram, H.; Grandolfo, M.; Scaini, D.; Gelain, F.; Casalis, L.;
Prato, M.; Giugliano, M.; Ballerini, L. Carbon nanotubes might improve
neuronal performance by favouring electrical shortcuts. Nature
Nanotechnology 2009, 4, 126-133.
(4) Lodish, H.; Berk, A.; Kaiser, C. A.; Krieger, M.; Scott, M. P.; Bretscher, A.;
Ploegh, H.; Matsudaira, P. Molecular Cell Biology; 6° ed.; W. H. Freeman, 2007.
Conclusion
95
Conclusion
In this thesis we have pursued biological applications of functionalized
CNTs.
As we have already stated, among the many applications of CNTs, the
biological ones are the most criticized. The debate mainly concerns the toxicity and
the uncertain fate of the CNTs inside the human body. Nevertheless, by dendrimer
functionalization of CNTs we have obtained evidence that make us confident in a
possible future exploitation of functionalized CNTs in drug delivery. We obtained
successful results in gene silencing with our dedrimeric compounds and the
compatibility of CNT-coated glasses with neuronal cultures was clearly proved,
demonstrating the reduced toxicity of these materials. Since neuronal cells are
very sensitive to contaminants and their viability is an important parameter to
assess the hazard of exposure to possible toxic agent, we can consider these results
rather comforting.
More in depth, for what concerns drug delivery, we can state that CNTs have
crucial properties, such as the high surface area and the possibility to penetrate the
cell membranes as nano-needles, that make them very promising for drug delivery
application. With respect to the choice of covalent functionalization, 1,3-dipolar
cycloaddition reaction of azomethine ylides is one of the most used and suitable.
Together with defect functionalization, good results have been obtained, making
possible to attach to the CNT dendrimeric-like structure and eventually add
different molecules creating a great opportunity in this field that badly needs new
types of efficient drug carriers. In our group, we are strongly pushing the direction
to obtain derivatives with enhanced bioavailability and compatibility, which could
be successfully adapted to different types of drugs. To address this issue, multiple
functionalization of CNTs is needed and is currently under study in our laboratory.
Conclusion
96
On the other side, the main difficulty is represented by our lack of
knowledge about the fate of CNTs inside the body, and we are also actively
investigating this aspect. As far as we know, not surprisingly, there is no evidence
about a possible “digestion” of CNTs inside the cells or at a systemic level and
irrefutable evidence of exocytosis has not yet been obtained. Biodistribution
studies have shown that renal excretion is the preferential way of elimination, but
this has not been demonstrated for all the CNT derivatives, as mentioned in this
thesis. When we combine CNTs with dendrimers or polymers, the biological tests
become more and more complex. As a consequence, the fate of the compounds,
their biological activity and their toxicity will be difficult to assess.
The sharp shape of CNTs leads to many worries because they are compared
with asbestos fibers. Luckily this is not true, because CNTs employed in biological
studies are usually short (around 500 nm) and functionalized, instead asbestos
fibers are long and thick, and they easily injure the tissues provoking
mesothelioma in lungs. In parallel, a still small but possibly increasing research
field is represented by other carbon nanostructures like carbon nanodiamonds
(spherical carbon particles ranging from few nm to 100 nm in diameter), that have
been functionalized in a way that is similar to CNTs and have high surface area but
with a round shape that should not cause any inflammatory response. Even in this
case, though, the systemic fate remains so far unclear. With respect to other carbon
nanostructures, carbon nanohorns are also under study but they bear an indented
structure that may inflame tissues.
The main problem of nanotechnology, and especially of nanomedicine, is
that the used materials are completely new and there has been just little time to
explored their possible advantages and disadvantages in all their facets, like
chronic toxicity.
In contrast to this “difficult” scenario, we are instead strongly convinced
that CNT-coated substrates for neuronal cells growth represent already now great
frontier for neurobiology and pathology because the boosting of the neuronal
network activity is critical for people with an injured spinal cord. Further
Conclusion
97
investigations in this field are essential, and we think that a lot can be done in
principle. CNTs are conductive needles that pinch the neuronal cell membrane and
probably induce a short-cut that is crucial to determine the boosting of the entire
neuronal network activity. The details of this interaction are not completely clear
but both neurons and CNTs are able to carry charges, neurons using ions and CNTs
using electrons. Although, we are currently trying to pass from bidimensional
neuronal networks, to more complicated biological environments, towards the in
vivo final tests, we already have many pieces of evidence that make us so
optimistic.
In conclusion, even though a lot of studies, especially of the in vivo type, are still
missing to draw a complete picture of the CNT use in biology and medicine, we are
quite confident that in the future biological applications will be vigorously
explored and many improvements will be achieved for all carbon nanostructures.
Conclusion
98
Appendix A: Review of the experimental methods
99
Appendix A:
Review of the
experimental methods
used in this thesis
In this chapter we review all the methods employed to produce and
characterize our derivatives.
A.1 Carbon Nanostructures Functionalization using 1,3-dipolar
cycloaddition reaction
The 1,3-dipolar cycloaddition of azomethine ylides to fullerene was
reported for the first time in 19931. The azomethine ylide cycloaddition, yielding
fulleropyrrolidines1-3, contributed significantly to the development of new
nanocarbon-based materials for strategic fields such as solar energy conversion
and molecular medicine4-6. Therefore it represents a powerful tool to introduce
covalent functionalizations to CNTs7,8, either MWCNTs and SWCNTs (Scheme 1).
Exohedral addition of organic groups helps to control solubility and morphology
in the solid state and the electronic properties of the fullerene core and CNTs. In
Appendix A: Review of the experimental methods
100
this thesis, we extensively used 1,3-dipolar cycloaddition reaction as the
starting point in every synthesis.
Scheme 1. CNTs addition of azomethine ylides
The reaction employs an -aminoacid and either an aldehyde or a ketone.
As seen in Scheme 1, it occurs by the condensation of the amine in the aminoacid
and the carbonyl group to form a dipole intermediate, which is able to react with
the numerous double bonds present in the nanotubes, both on their sidewall and
on their tip. A well-established protocol consists on the addition of both reactants
in excess in DMF for 5 days at 130°C; in the presence of protecting functional
groups (like tert-butyloxycarbonyl, cleavable by means of high temperature) the
reaction temperature could be decreased to 115°C. To perform this reaction, we
have used {2-[2-(2-tert-butoxycarbonylamino-ethoxy)-ethoxy]-ethylamino}-acetic
acid reported in Scheme 2. By removal of the protective groups, several different
molecules can be attached to CNTs8. The presence of amino groups can be
assessed using a quantitative test used in solid phase synthesis, known as Kaiser
Test9,10. Side characterization techniques, as Kaiser Test, are in fact crucial to study
our derivatives, since the conventional chemical characterization tools, as NMR or
mass spectrometry, cannot be used with functionalized CNTs. 1H NMR analyses of
functionalized CNTs led to weak and broad signals that are difficult to univocally
Appendix A: Review of the experimental methods
101
interpret, even though new methodologies are arising in this direction, like
diffusion NMR11.
Scheme 2. 1) Synthesis of N-(tert-butyloxycarbonyl)-amino-diethoxy-ethyl amine. 2)Synthesis of {2-[2-(2-
tert-butoxycarbonylamino-ethoxy)-ethoxy]-ethylamino}-benzyl ester. 3)Synthesis of {2-[2-(2- tert-
butoxycarbonylamino-ethoxy)-ethoxy]-ethylamino}-acetic acid.
In chapter 2 we reported a well-constructed synthesis of positively charged
PAMAM-modified-CNTs derivatives12. The use of dendrimer-like structures fulfills
the main requirements of covalent carbon nanotubes chemistry. Exploiting 1,3-
dipolar cycloaddition and the growing of a dendron, it is possible to provide an
increased number of reactive groups at the periphery without damaging the
carbon nanostructure.
In fact, dendrimers are complex polymers and own a peculiar branched-
architecture with tunable nanoscale dimensions: they are made of an internal core,
and interior layers, composed of repeating units named generations (indicated as
G), and an exterior layer that bears the peripheral functionalities (Scheme 3).
1)
2)
3)
Appendix A: Review of the experimental methods
102
Scheme 3. (A) Dendrimer and (B) Dendron structures, generation 0 (G0) is indicated in black, generation
1 (G1) in blue, and generation 2 (G2) in red.
In Scheme 3 is indicated a derivative of dendrimer family, that is dendron.
Dendron is grown developing a focal point only towards one part of the plain; by
means of this branched polymeric dendron, we could functionalize CNTs
introducing many amines.
A.2 Microwave
MW can be exploited to perform organic reactions. Their use as a powerful heating
source in organic synthesis arose in the middle of 1980s. From the beginning, this
approach has received wide attention among organic chemists because it can
provide high yield, reduce side reactions, improve reproducibility and, above all,
reduce chemical reaction times, of order of magnitudes13. In chapter 2, we
proposed to use microwave for 1,3-dipolar cycloaddition with ionic liquids. This
combination lead us to obtain one the highest yield reported (entry 14 Table 2
Chapter 2)14.
Microwave frequencies range from 0.3 to 300 GHz. Domestic ovens and
reactors for chemical reactions typically operate at 2.45 GHz (12.24 cm
wavelength) to avoid interference with telecommunication and cellular phone
frequencies. The correspondent energy is of 0.0016 eV (0.037 Kcal/mol), too small
to induce chemical reactions, as it is possible to see from the comparison in Table
115.
A
B
Appendix A: Review of the experimental methods
103
Table 1. Brownian motion and bonds energies.
Energy (eV) Energy
(Kcal/mol)
Brownian Motion 0.025 (298K) 0.56
Hydrogen Bonds 0.043 to 0.65 1 to 15
Covalent Carbon Bonds 1.73 to 5.204 40 to 120
MW induces a “dielectric heating” of the sample, based on the conversion of
electromagnetic energy absorbed by the medium into heat. The MW electric field
induces a polarization of the molecules at the same frequency of the light. In the
infrared regime, electromagnetic field induces atomic vibrations in molecules and
crystals and polarization processes resulting from the dipole moment induced by
distorsion of nuclei position. In the presence of an electric field a dipole moment
undergoes a torque that tends to orientate it parallel to the electric field,
dissipating energy into the medium. The minimum in potential energy is when the
dipole moment takes the same direction of the electric field. The electric field
vector switches its orientation approximately every picosecond: the result is that
the torque induces rotation of polar molecules that cannot always stand this rate.
So, the delay between electromagnetic stimulation and molecular response is
defined as dielectric loss (휀′′) and it expresses the efficiency to convert
electromagnetic energy into heat. Defining the dielectric constant (휀′) as the ability
of molecules to be polarized by the electric field, the quotient tan 𝛿 is derived at
(1):
tan 𝛿 = 휀′′ 휀 ′ (1)
Higher tan 𝛿, higher the efficiency of microwave absorption: this value is often
used to classify microwave absorbing property of organic solvents (tan 𝛿> 0.5 high;
tan 𝛿= 0.1-0.5 medium; tan 𝛿< 0.1 low). An important note about CNTs is that they
have a metal content because of the catalyst particle residue, made mainly of metal
oxides such as iron, nickel and cobalt. These metal oxides show very high magnetic
Appendix A: Review of the experimental methods
104
losses that occur in MW region, and consequently they are optimal absorbers of
MW. On the other hand, when the dielectric loss of solvents or reactants is too
small, the addition of such metals improves dielectric heating. However,
thermodynamic effects of electric field over chemical equilibrium are very well
known: when the product has a large dipole moment in respect with reactants the
equilibrium is shifted towards the product under the action of an applied electric
field13,16. Besides this, the conventional thermal heating method uses an external
heat source, such as in oil bath. Then transferring energy into the system is not so
efficient and there is a gradient of temperature towards the reaction mixture,
where the temperature is lower. On the other hand, dielectric heating implies that
the reaction mixture is directly heated and the gradient of temperature is inverted
and lower towards the outsides. To perform microwave reaction, vessels
transparent to this wavelength are employed, such as borosilicate glass, quartz or
Teflon. To this aim, quartz is particularly resistant and suitable for pristine carbon
nanotubes reaction, whereas, we have found that borosilicate glass can be used
with purified carbon nanotube (Chapter 2).
Brunetti et al17 synthesized SWCNT derivatives with the combination of 1,3
dipolar cycloaddition and addition of diazonium salts, using doses of microwave
higher than in our case. In general, the use of MWs with CNTs is still under
development, and many different strategies can be in principle tested.
A.3 Thermogravimetric Analysis and Kaiser Test
TGA is one of the most important analytical techniques used to determine
the quantity of functional groups attached on fucntionalized CNT surface. It is
based on the thermal stability of materials: the change of weight of the substance is
monitored as a function of the increasing temperature, with sensitivity in the
nanogram range. All the derivatives described in this thesis were characterized by
means of TGA.
Appendix A: Review of the experimental methods
105
TGA can be performed in different environments, optimized according to
the experiment needs: an inert atmosphere (nitrogen or argon) is chosen for
measuring the content of organic moieties bound to CNTs, which start burning at
about 350°C; an oxygen atmosphere is instead chosen to evaluate the metal
residue content (such as the metal catalysist used in the synthesis or the metal
nanoparticles subsequently attached), which are converted into solid oxides after
the burning of CNTs at about 500°C. The system sensitivity is given by the
precision of the weight balance and of the power supply which provides the
temperature ramp of the furnace. We usually adopt a linear temperature ramp
(10°C/min) preceded by an equilibration step in which the sample is annealed at
around 100°C for 20 minutes. We ramp the temperature up to 1000°C, looking for
a good compromise between sensitivity and overall time. The main pieces of
information regard the weight % (ash content, i.e. residual mass) and the
temperature at which the organic volatile moieties burn (oxidation temperature),
simply using derivative functions. To our purpose, which is the determination of
the degree of functionalization of CNT substrates, we proceed in a differential
manner: we evaluate the percentage loss of weight of the system for a family of
different derivates grown step by step starting from the same progenitor, to
correlate the number of functional groups to the number of carbon atoms (2)
and/or to determine the concentration of the functional groups in the carbon
matrix (3):
𝐶𝑎𝑟𝑏𝑜𝑛 𝑎𝑡𝑜𝑚𝑠𝐹𝑢𝑛𝑐𝑡 𝐺𝑟 =
𝑊𝑒𝑖𝑔𝑡 % × 𝑀𝑜𝑙𝑒𝑐𝑢𝑙𝑎𝑟 𝑤𝑒𝑖𝑔𝑡 𝑜𝑓 𝑓𝑢𝑛𝑐𝑡 𝑔𝑟
𝑀𝑜𝑙𝑒𝑐𝑢𝑙𝑎𝑟 𝑤𝑒𝑖𝑔𝑡 𝑜𝑓 𝐶 × 𝐿𝑜𝑠𝑠 𝑜𝑓 𝑤𝑒𝑖𝑔𝑡 % (2)
𝑚𝑚𝑜𝑙
𝑔=
𝐿𝑜𝑠𝑠 𝑜𝑓 𝑤𝑒𝑖𝑔𝑡 % × 10
𝑀𝑜𝑙𝑒𝑐𝑢𝑟𝑎𝑟 𝑊𝑒𝑖𝑔𝑡 𝑜𝑓 𝑓𝑢𝑛𝑐𝑡𝑖𝑜𝑛𝑎𝑙 𝑔𝑟𝑜𝑢𝑝 3
Where 𝑓𝑢𝑛𝑐𝑡 𝑔𝑟 means functional groups, and 𝑀𝑜𝑙𝑒𝑐𝑢𝑙𝑎𝑟 𝑤𝑒𝑖𝑔𝑡 𝑜𝑓 𝐶 is equal to
12.
Appendix A: Review of the experimental methods
106
The results obtained by TGA are then compared to Kaiser Test outputs. This
test, introduced in 19709 is a quantitative measurement of free amino groups in
solid state peptide synthesis. It is crucial to evaluate the success of 1,3 dipolar
cycloaddition reaction occured with the aminoacid reported in Scheme 2. As
already mentioned, the main goal of the synthesis is to succeed in the formation of
free amino groups released after deprotection. Kaiser Test is a simple ninhydrin
test, which follows the reaction shown in scheme 4:
Scheme 4. Kaiser test reaction. Complex a is responsible for absorption at 570 nm.
This reaction, which takes place at 100°C, produces a ninhydrin complex
(complex a in the scheme) which absorbs light at 570 nm. By means of the
recorded absorbance, we are able to quantify the free amino groups: according to
the Lambert-Beer law, absorbance can be in fact correlated with the concentration
of amino groups as follows:
𝑚𝑚𝑜𝑙
𝑔=
𝐴𝑏𝑠 𝑠𝑎𝑚𝑝𝑙𝑒 − 𝐴𝑏𝑠 𝑏𝑙𝑎𝑛𝑘 × 𝑑𝑖𝑙𝑢𝑡𝑖𝑜𝑛 𝑚𝐿 × 103
𝐸𝑥𝑡𝑖𝑛𝑐𝑡𝑖𝑜𝑛 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 × 𝑆𝑎𝑚𝑝𝑙𝑒 𝑤𝑒𝑖𝑔𝑡 𝑚𝑔 4
Where the 𝐴𝑏𝑠 𝑠𝑎𝑚𝑝𝑙𝑒 is the absorbance of resulting complex (Scheme 4);
𝐴𝑏𝑠 𝑏𝑙𝑎𝑛𝑘 is referred to the absorbance of the blank (the mixture of ethanolic
solution of phenol, KCN in pyridine and ethanolic solution of ninhydrine) at 570
nm; 𝑑𝑖𝑙𝑢𝑡𝑖𝑜𝑛 is related to the reaction process (the sample to be analyzed at the
spectrophotometer is diluted of 5 mL); 𝐸𝑥𝑡𝑖𝑛𝑐𝑡𝑖𝑜𝑛 𝑐𝑜𝑒𝑓𝑓𝑖𝑐𝑖𝑒𝑛𝑡 is 15000 m-1 cm-1;
Appendix A: Review of the experimental methods
107
𝑆𝑎𝑚𝑝𝑙𝑒 𝑤𝑒𝑖𝑔𝑡 is referred to the weight of CNTs used in the analysis. The
sensitivity reported for this test is up to 5 mol/g9.
In general, a good estimation of the functional groups content in our CNTs is
consistently obtained by means of these two complementary techniques.
A.4 Raman Spectroscopy
Raman scattering, based on the inelastic scattering of monochromatic light, is a
valuable technique to study vibrational and rotational excitations of a molecular
system. We have use Raman analyses to obtain evidence of covalent (Chapter 2)
functionalization of CNTs.
When monochromatic light beam (typically a visible laser beam) hits a
sample, a small fraction of it is elastically scattered (Rayleigh scattering) and a
smaller fraction (one photon over 107) is inelastically scattered (Raman
scattering). In Raman scattering the difference in energy between incident and
inelastically scattered photons corresponds to the vibrational frequencies of the
scattering molecules. Regarding the incident light, the scattered beamcan exibit
lower or higher frequencies, known as Stokes and Anti-Stokes scattering
respectively (Figure 1).
Figure 1. Schematic representation of elastic (Rayleigh) and inelastic (right: Stokes, left: anti-Stokes),
adapted from reference18.
Appendix A: Review of the experimental methods
108
Stokes radiation implies the energy transfer from the photon to the molecule, and
the photon is consequently scattered at lower energy; the Anti-Stokes radiation is
due to the energy transfer from the molecule to the photon, and the resulting
radiation appears at higher energy and lower wavelength. Among the two, the
Stokes process is stronger, since at room temperature the population state of a
molecule is essentially in its ground vibrational state.
The quantity measured in Raman spectroscopy is the Raman shift, given by (5)
𝑅𝑎𝑚𝑎𝑛 𝑠𝑖𝑓𝑡 = 𝜈 ± 𝑣 𝑚 (5)
Where 𝜈 is the wavenumber (cm-1) of the incident radiation and 𝑣 𝑚 is the
wavenumber (cm-1) of the energy difference between the lowest and first excited
vibrational energy levels.
The choice of the wavelength of the excitation laser is critical: it should be chosen
optimizing mainly i) Raman scattering cross section (which is proportional to the
inverse of the fourth power of the wavelength, therefore calling for lower
wavelengths) ii) magnitude of the Raman shift in wavelength (see (5)) and iii) the
potential for activating fluorescence (which is higher at lower wavelengths)18.
A Raman set-up is made of a monochromator, adequate filters to allow the
detection at very low Raman shift, and a detector that usually is a charge-coupled
device detector (CCD).
Optical and spectroscopic properties of carbon nanotubes are mainly
characterized by the one-dimensional confinement of the electronic and phonon
structure which is more relevant for single walled CNTs. Raman phonons involve
normal vibration modes, that play a key role as carrier of thermal energy in
conduction processes and in determining thermodynamic properties. Raman
scattering depends on specific phonon modes. To have a better understanding of
CNTs Raman signals, we will use a comparison with Raman spectrum of
graphene/graphite. The electronic bands are responsible for the strong in-plane
covalent bond within the graphene sheets, whereas the bands are responsible for
Appendix A: Review of the experimental methods
109
van der Waals interactions between graphene sheets in 3D graphite. Furthermore,
bands are close to Fermi level and electrons can be excited from valence band ()
to the conduction band (*) optically. During the scattering event, it is possible to
distinguish different steps: first an electron is excited from the valence to the
conduction band, then this electron is scattered by emitting or absorbing phonons,
finally the electron relaxes to valence band emitting a photon. The number of
emitted phonons can vary and it could be one-phonon, two-phonon or multi-
phonon Raman process. The number in the sequence of total scattering events,
including elastic scattering by an imperfection, is called “order of a scattering
event”.
G (graphite) band and RBM (radial breathing mode) are first-order scattering
event, involving one phonon emission. In SWCNTs, G band splits in many
frequencies around 1580 cm-1 and it is related to the conductive properties of the
tube, but in general it is the more intense band, to which the spectrum is usually
normalized. The G peak corresponds to the graphite E2g optic mode, that is the
doubly degenerate in-plane optical vibration and that is Raman active (Figure 2).
Figure 2. E2g vibration mode of graphene.
On the other hand, RBM is a bond-scattering out of plane phonon mode and
frequencies are about 100-500 cm-1. The RBM frequency is inversely proportional
to the tube diameter and the mass of all the carbon atoms along the
circumferential direction is proportional to the diameter.
Second order scattering, instead, consists of two-phonon scattering events or one-
phonon and one elastic scattering event: a typical example is the D-band at 1350
Appendix A: Review of the experimental methods
110
cm-1, related to sp3 carbon atoms in carbon nanotubes. D band is the so-called
“disorder” band and it may come from a symmetry-lowering effect: it is due to
defects or nanotube caps, bending of CNTs or presence of carbon nanoparticles
and/or amorphous carbon. In graphite, the intensity of D band compared with the
Raman-allowed E2g mode depends on the size of graphite microcrystals and
consequently on the order of the sample. Comparing spectra of different
derivatives, it is possible to identify whether a derivative has been functionalized
compared to the pristine material. This feature, indeed, is the main useful
measurement for our purpose.
Another common feature in a carbon nanotube Raman spectra is the G’ band, a
second order scattering two-phonon mode, with a frequency of 2700 cm-1. This
overtone is associated with the over-bending of the longitudinal optical branches
of graphite19,20.
Generally, in the literature, there are many examples in which Raman is
used to study CNT properties21. As already mentioned, RBM are an important
feature to discriminate metallic and semiconducting CNTs22. Moreover, interesting
studies have been recently published by Dai et al23 about the use of SWCNTs in
Raman imaging of biological system.
A.5 Visible-Near Infrared Spectroscopy
UltraViolet-Visible-Near Infrared Spectroscopy (UV-Vis-NIR) investigates
SWCNTs in solution and can distinguish between metallic and semiconducting
ones. We have used this technique to have a rough idea of the degree of
functionalization of our SWCNT-derivative (Chapter 2).
The UV-Vis-NIR absorption is based on a peculiar characteristic of CNTs: the
fact that they essentially are one-dimensional systems. As a consequence,
electronic density of states consists of a series of Van Hove singularities and
absorption spectra are representative of transitions between them. Simply, the
density of states (DOS) are the number of one-electron levels at each energy that
Appendix A: Review of the experimental methods
111
can be occupied. In 1d system, like CNT, the DOS is characterized by some
divergences, which are called Van Hove singularity. This measurement is
performed in solution to obtained the response of individual SWCNTs of different
chiral indices and/or of different diameter. Three broad bands may be observed:
one is due to metallic (M1=600-700 nm), and the other two are due to
semiconducting (S1=1600-1700 nm and S2=1000 nm). The outcome of the
spectrum depends on different parameters, as solubility, concentration and bundle
formation and it is sometimes very difficult to be reproducible; but what is more
important is to demonstrate that we have introduced functional groups. In this
case, the one-dimensional system becomes at least a two-dimensional system and
absorption spectrum changes between not-functionalized and functionalized CNTs,
because van Hove singularities band are lost. This depletion of bands permits a
qualitative analysis of the degree of functionalization and it is important to notice
is the decrease of intensity in absorption spectra between the starting material and
the functionalized one22,24.
As mentioned, UV-Vis-NIR spectroscopy can be exploit to distinguish
between semiconducting and metallic nanotubes: Campidelli et al reported an
attempt of separation via covalent functionalization exploiting NIR (and Raman) as
a characterization tool22.
A.6 Trasmission and Scanning Electron Microscopy
Electron microscopy (EM) is a powerful technique to visualize
nanostructures; this is why it represents a valuable method used to characterize
CNT derivatives. TEM is constantly employed in the whole thesis, while SEM
(Chapter 4) is restricted to those derivatives deposited on substrates.
Appendix A: Review of the experimental methods
112
The introduction of EM is strictly related to the scientific revolution
between the end of the XIX and the beginning of XX century, after electrons were
discovered in 1897 by J.J. Thompson. E. Ruska, Nobel laureate in 1986 with G.
Binnig and H. Rohrer, has significantly contributed to the description of the first
electron microscope in 1934 and then to the first prototype of transmission
electron microscope, realized in 1938 for “Siemens&Halske AG” company
improving the Ruska’s original microscope. In principle, an electron microscope is
quite similar to a light microscope but with a resolution thousand times higher.
According to the de Broglie principle in fact, electrons have a wavelength that is
inversely proportional to their energy. In electron microscopes, electrons are
extracted from a hot filament, and accelerated to energies up to few hundreds of
thousands V, correspondent to wavelength down to tenths of nanometers. The
main part of electron microscopes is given by the column, which host the electron
gun, the acceleration field, and a series of focusing lens (condensers or magnetic
fields lens) and apertures, that control the size of the beam. The microscope
operate under vacuum. At the beam/specimen interface several types of signals
are generated due to the complex interactions of electrons with matter, detected
and processed to produce an image.
Transmission Electron Microscope25 (TEM) is based on the detection of high
electron energy transmitted through a very thin sample (<100 nm thick). In the
bright-field imaging mode, the contrast is given by the different absorption of
different portions of the sample. In the case of crystalline samples, electrons are
diffracted and diffraction peaks positions are detected to give a contrast image. A
pattern of diffraction dots is obtained in the case of perfectly crystalline materials,
while a series of ring is observed for polycrystalline or amorphous materials.
HiRes TEM is an alternative working mode for TEM. In this case, differences in
phase of the electrons that have interacted with the sample are detected. The
HiRes mode highlights the structure of nanometer-size objects, as nanoparticles or
CNTs. There is also the possibility to perform spectroscopy like Electron Energy
Appendix A: Review of the experimental methods
113
Loss Spectroscopy (EELS) and Energy Dispersive X-Ray (EDX) analysis, known also
as micro-analysis. Samples are deposited on grids from a solution. The grids used
for CNTs, and also for many other nanoparticles, are made of a metallic framework
(Ni or Cu) and covered with a carbon film, continuous or with holes, thin enough to
transmit electrons. When the sample cannot be solubilized, it is necessary to make
an inclusion with resins and then cut thin film to deposit on the grid.
TEM is widely used for CNTs analysis, in order to investigate the dispersion and
the cleanness of functionalized nanotubes, while giving essential morphological
information. In our case, we have used TEM (Chapter 2, 3) to characterized
reported compounds.
The SEM26 (SEM), instead, is a fundamental technique to image the surface
topography of micro and nano structured samples without any requirements for
sample preparation. A focused beam is scanned over the surface and the emitted
electrons derived by the interactions with the sample are detected. Three imaging
signals can be interaction volume, back scattered electrons (BSE) and secondary
electrons (SE). Interaction volume depends on the beam energy and the atomic
number of atoms of the specimen. BSE are incident beam electrons that escape
from the surface without any significant change in energy. BSE analysis is
particularly interesting when information about composition is needed: it depends
on chemical composition and surface morphology of samples with different atomic
composition (i.e. with diverse atomic numbers), it is possible to highlight
differences in contrast among the surface because the intensity of BSE depend on
the second power of the atomic number.
SE, instead, depend only on surface morphology and are due to inelastic
scattering. They are weakly bound outer shell electrons that receive sufficient
energy to be ejected from the atoms. SE can be generated directly by the beam or
by BSE and their dependence on the atomic number is negligible. While BSE return
both compositional and morphological information, SE give back morphological
Appendix A: Review of the experimental methods
114
information because the surface sensitivity is higher. Also with SEM it is possible to
perform EDX analysis.
A.7 Atomic Force Microscopy
The Atomic force microscope (AFM), invented by Binnig, Quate and Gerber
in 198627,28, is part of the Scanning Probe Microscopy (SPM) family: it is a versatile
technique that senses forces (Pauli, Van der Waals, electrostatic, etc.) between an
extremely small tip and a sample. It can measure tip-sample interactions without
any restriction over the conductivity of the sample (dissimilarly from STM). The
specimen seen by AFM does not need then to be flat, a limitation that instead apply
to STM measurements: the surface in STM must be flat because the tip-sample
separation is monitored as a function of the flow of electrons and this separation is
typically about 1 nm.
We used to measure the height profile of covalent functionalized MWCNTs
(Chapter 2), in non-contact mode. CNTs were deposited on a flat surface of mica by
spin coating deposition.
In AFM the probe is a sharp tip, placed at the end of a flexible cantilever, and
a laser beam is focused on the back of it (cantilevers can be metal coated to
improve the laser reflectivity). The reflected laser beam hits a four quadrants
Position Sensitive Photo Diode (PSPD) that is used to track the laser spot
movements induced by cantilever bending and torsion while the tip is scanning the
surface. In this way, using a feedback mechanism it is possible to maintain constant
the force applied by the tip on the surface (contact mode or, more properly, static
mode) or the amplitude (non-contact mode or dynamic mode). Cantilevers are
flexible and they are comparable to springs, those are described by Hook’s law. By
means of this mathematical relation, geometric calculations about the laser
displacement over the PSPD return information about occurring interactions.
Appendix A: Review of the experimental methods
115
Figure 3. Schematic representation of AFM system.
The feedback system controls tip-specimen interactions, using a setpoint that is
mantained while measuring. There are two ways to define a setpoint, one for
contact, the other for non-contact mode: in contact mode the set-point is
determined by the cantilever deflection (higher the cantilever deflection, higher
the force applied and, consequently, larger tip-surface interaction; lower is the
force, weaker the tip-surface interaction), while in non-contact mode it is
determined by the variations in the cantilever resonance frequency (FM) or
amplitude at a fixed frequency (AM). Generally, in non-contact mode, the
interaction with the specimen is smaller than in contact mode, consequently
dynamic mode is usually the preferred technique to image soft samples. Tip and
sample movements are extremely precise and are driven by a ceramic piezo
material, in a tube scanner shape: the scanner is able to perform sub-nanometric
movements.
AFM-tip can also be used as a tool to immobilize DNA molecules in order to
produce both DNA29 or protein chip30, by means of nanografting technique.
Appendix A: Review of the experimental methods
116
Notwithstanding, topographic AFM images of neuronal cells grown on a carpet of
MWCNTs have been also reported31.
A.7.1 Electrostatic Force Microscopy
Electrostatic Force Microscopy32,33 (EFM) is an AFM working mode, in
which a biased tip is scanned over a surface to probe surface potential and charge
distribution. We employed this technique to characterize positively charged
functionalized CNTs (Chapter 2), exploring the possibility to draw the CNT profile
highlighting the presence of positive charges.
The working principle is based on the Coulomb interactions occurring
between the sample and the tip, and it causes the subsequent changes in the
oscillation amplitude and phase of the AFM cantilever vibrating at a fixed
frequency. The different power law between van der Waals and Coulomb forces
(van der Waals forces depend on the sixth power of the distance, while Coulomb
forces depend on the second power) allows the discrimination between
Coulombian interaction due to charge accumulation, and van der Waals
interaction, responsible of the morphology connected forces. EFM differs from
Kelvin Probe Microscope (KPM), that gives information about the surface potential
of the sample, because the substrate may not be conductive, whereas the tip is a
metal-coated tip, and, moreover, EFM is generally characterized by higher spatial
resolution.
In EFM, to separate van der Waals and Coulomb contributes two techniques
can be used: the most commonly adopted one is the second-pass technique. The
second-pass technique, coherently with the van der Waals law, draws the
topographic profile in the first pass, then the tip is lifted to a fixed height (usually
50 nm); in the second pass changes in amplitude and phase of the tip oscillations
(mainly due to the Coulomb contribution) are measured giving information about
surface charge distribution. The drawback in the tip lift is a partially loss in image
resolution. The second technique is the Enhanced-EFM (E-EFM) able to perform in
Appendix A: Review of the experimental methods
117
one single pass both the topography and the electrostatic measurements because
of the presence of an external lock-in, able to separate the two force contributions
due to the different frequency dependence. In our E-EFM set-up, a conductive tip
scans a surface in close proximity in non-contact mode, when a DC bias is applied
to the cantilever. An external lock-in amplifier applies simultaneously an AC
voltage, in addition to the DC bias applied by the AFM controller, allowing for
discriminating between electrostatic and van der Waals contributions to the
measured force34. Changing the bias of the tip, different contrast and phase and
amplitude values are measured: there are several discussion about the meaning of
the recorded signals but the phase shift, that exists between the driving sinusoidal
signal and the cantilever response (read by the PSPD), allows to identify whether
the operating regime is attractive or repulsive. Phase-shift measurements provide
a practical and unambiguous method to identify the operating regime: conditions
that imply positive phase shifts are characteristic of an attractive regime, while
conditions that set negative phase shifts are characteristic of a repulsive regime.
The phase shift is proportional to the total charge present below the tip.
We tried to use this tool to investigate the presence of positive charges on
MWCNTs 8 (Chapter2), because of the results obtained by Navarro-Gomez et al35:
in their study they compared the different conductivity of a DNA nanowire and of a
SWCNT. Later on, other studied on the electrostatic response of bare nanotubes36
were reported, but our expertise in functionalization should allow us to investigate
the presence of positive charges on the nanotubes surface.
References:
(1) Maggini, M.; Scorrano, G.; Prato, M. Addition of azomethine ylides to C60:
synthesis, characterization, and functionalization of fullerene pyrrolidines.
Journal of the American Chemical Society 1993, 115, 9798-9799.
(2) Prato, M.; Maggini, M. Fulleropyrrolidines: A Family of Full-Fledged Fullerene
Derivatives. Accounts of Chemical Research 1998, 31, 519-526.
Appendix A: Review of the experimental methods
118
(3) Tagmatarchis, N.; Prato, M. The addition of azomethine ylides to [60]fullerene
leading to fulleropyrrolidines. Synlett 2003, 768-779.
(4) Prato, M. [60]Fullerene chemistry for materials science applications. J. Mater.
Chem. 1997, 7, 1097-1109.
(5) Mateo-Alonso, A.; Guldi, D.; Paolucci, F.; Prato, M. Fullerenes: Multitask
Components in Molecular Machinery. Angewandte Chemie International
Edition 2007, 46, 8120-8126.
(6) Mateo-Alonso, A.; Sooambar, C.; Prato, M. Synthesis and applications of
amphiphilic fulleropyrrolidine derivatives. Org. Biomol. Chem. 2006, 4, 1629-
1637.
(7) Tasis, D.; Tagmatarchis, N.; Bianco, A.; Prato, M. Chemistry of carbon
nanotubes. Chemical Reviews 2006, 106, 1105-1136.
(8) Singh, P.; Campidelli, S.; Giordani, S.; Bonifazi, D.; Bianco, A.; Prato, M. Organic
functionalisation and characterisation of single-walled carbon nanotubes.
Chemical Society Reviews 2009, 38, 2214-2230.
(9) Kaiser, E.; Colescott, R. L.; Bossinger, C. D.; Cook, P. I. Color test for detection
of free terminal amino groups in the solid-phase synthesis of peptides.
Analytical Biochemistry 1970, 34, 595-598.
(10) Sarin, V.; Kent, S.; Tam, J.; Merrifield, R. Quantitative monitoring of solid-
phase peptide synthesis by the ninhydrin reaction. Analytical Biochemistry
1981, 117, 147-157.
(11) Marega, R.; Aroulmoji, V.; Dinon, F.; Vaccari, L.; Giordani, S.; Bianco, A.;
Murano, E.; Prato, M. Diffusion-Ordered NMR Spectroscopy in the Structural
Characterization of Functionalized Carbon Nanotubes. Journal of the
American Chemical Society 2009, 131, 9086-9093.
(12) Herrero, M. A.; Toma, F. M.; Al-Jamal, K. T.; Kostarelos, K.; Bianco, A.; Da Ros,
T.; Bano, F.; Casalis, L.; Scoles, G.; Prato, M. Synthesis and Characterization of
a Carbon Nanotube−Dendron Series for Efficient siRNA Delivery. Journal of
the American Chemical Society 2009, 131, 9843-9848.
Appendix A: Review of the experimental methods
119
(13) Loupy, A. Microwaves in Organic Synthesis; 2° ed.; Wiley-VCH, 2006.
(14) Guryanov I.; Toma F. M.; Montellano López A.; M., C.; Da Ros T.; Angelini G.;
D'Aurizio E.; Fontana A.; Prato M.; Bonchio M. Microwave-Assisted
Functionalization of Carbon Nanostructures in Ionic Liquids. Chemistry - A
European Journal.
(15) Goshe, A. J.; Steele, I. M.; Ceccarelli, C.; Rheingold, A. L.; Bosnich, B.
Supramolecular recognition: On the kinetic lability of thermodynamically
stable host–guest association complexes. Proceedings of the National
Academy of Sciences of the United States of America 2002, 99, 4823-4829.
(16) Kappe, C. O. Controlled Microwave Heating in Modern Organic Synthesis.
Angewandte Chemie International Edition 2004, 43, 6250-6284.
(17) Brunetti, F. G.; Herrero, M. A.; Munoz, J. D. M.; Diaz-Ortiz, A.; Alfonsi, J.;
Meneghetti, M.; Prato, M.; Vazquez, E. Microwave-Induced Multiple
Functionalization of Carbon Nanotubes. Journal of the American Chemical
Society 2008, 130, 8094-8100.
(18) Browne, W. R.; McGarvey, J. J. The Raman effect and its application to
electronic spectroscopies in metal-centered species: Techniques and
investigations in ground and excited states. Coordination Chemistry Reviews
2007, 251, 454-473.
(19) Dresselhaus, M.; Dresselhaus, G.; Saito, R.; Jorio, A. Raman spectroscopy of
carbon nanotubes. Physics Reports 2005, 409, 47-99.
(20) Saito, R.; Dresselhaus, G.; Dresselhaus, M. S. Physical Properties of Carbon
Nanotubes; 1° ed.; World Scientific Publishing Company, 1998.
(21) Jorio, A.; Dresselhaus, Gene; Dresselhaus, Mildred S. Carbon Nanotubes:
Advanced Topics in the Synthesis, Structure, Properties and Applications; 1° ed.;
Springer, 2008.
(22) Campidelli, S.; Meneghetti, M.; Prato, M. Separation of Metallic and
Semiconducting Single-Walled Carbon Nanotubes via Covalent
Functionalization. Small 2007, 3, 1672-1676.
Appendix A: Review of the experimental methods
120
(23) Liu, Z.; Li, X.; Tabakman, S. M.; Jiang, K.; Fan, S.; Dai, H. Multiplexed Multicolor
Raman Imaging of Live Cells with Isotopically Modified Single Walled Carbon
Nanotubes. Journal of the American Chemical Society 2008, 130, 13540-
13541.
(24) Krupke, R.; Hennrich, F.; Hampe, O.; Kappes, M. M. Near-Infrared Absorbance
of Single-Walled Carbon Nanotubes Dispersed in Dimethylformamide. The
Journal of Physical Chemistry B 2003, 107, 5667-5669.
(25) Williams, D. B.; Carter, C. B. Transmission Electron Microscopy: A Textbook for
Materials Science; 2° ed.; Springer, 2009.
(26) Goldstein, J.; Newbury, D. E.; Joy, D. C.; Lyman, C. E.; Echlin, P.; Lifshin, E.;
Sawyer, L.; Michael, J. Scanning Electron Microscopy and X-ray Microanalysis;
3° ed.; Springer, 2003.
(27) Smith, D.; Binnig, G.; Quate, C. Atomic point-contact imaging. Applied Physics
Letters 1986, 49, 1166-1168.
(28) Ohnesorge, F.; Binnig, G. True atomic resolution by atomic force microscopy
through repulsive and attractive forces. Science 1993, 260, 1451-1455.
(29) Castronovo, M.; Radovic, S.; Grunwald, C.; Casalis, L.; Morgante, M.; Scoles, G.
Control of Steric Hindrance on Restriction Enzyme Reactions with Surface-
Bound DNA Nanostructures. Nano Letters 2008, 8, 4140-4145.
(30) Bano, F.; Fruk, L.; Sanavio, B.; Glettenberg, M.; Casalis, L.; Niemeyer, C. M.;
Scoles, G. Toward Multiprotein Nanoarrays Using Nanografting and DNA
Directed Immobilization of Proteins. Nano Letters 2009, 9, 2614-2618.
(31) Cellot, G.; Cilia, E.; Cipollone, S.; Rancic, V.; Sucapane, A.; Giordani, S.;
Gambazzi, L.; Markram, H.; Grandolfo, M.; Scaini, D.; Gelain, F.; Casalis, L.;
Prato, M.; Giugliano, M.; Ballerini, L. Carbon nanotubes might improve
neuronal performance by favouring electrical shortcuts. Nature
Nanotechnology 2009, 4, 126-133.
Appendix A: Review of the experimental methods
121
(32) Martin, Y.; Williams, C. C.; Wickramasinghe, H. K. Atomic force microscope--
force mapping and profiling on a sub 100-[A-ring] scale. Journal of Applied
Physics 1987, 61, 4723-4729.
(33) Girard, P. Electrostatic force microscopy: principles and some applications to
semiconductors. Nanotechnology 2001, 12, 485-490.
(34) Park Systems Atomic Force Microscope, AFM/SPM.
(35) Gómez-Navarro, C.; Moreno-Herrero, F.; de Pablo, P. J.; Colchero, J.; Gómez-
Herrero, J.; Baró, A. M. Contactless experiments on individual DNA molecules
show no evidence for molecular wire behavior. Proceedings of the National
Academy of Sciences of the United States of America 2002, 99, 8484-8487.
(36) Wang, Z.; Zdrojek, M.; Melin, T.; Devel, M. Electric charge enhancements in
carbon nanotubes: Theory and experiments. Physical Review Letter B 2008,
78, 85425-85430.
Appendix A: Review of the experimental methods
122
Appendix B: experimental section
123
Appendix B:
experimental SECTION
B.1 Materials
MWCNTs were purchased from Nanostructured & Amorphous Materials Inc.,
Huston, TX; Stock #: 1240XH, 95%, OD 20-30 nm). HiPCO purified SWCNTs CNI®
were received from Carbon Nanotechnologies Incorporated. Chemicals and
solvents for synthesis were obtained from Sigma-Aldrich and they were used
without further purification. The radioactive tracer [111In]Cl3 was obtained from
Amersham Pharmacia Biosciences as an aqueous solution and used without
further purification.
Non-coding siRNA was obtained from Eurogentec. The proprietary
sequence siCONTROL TOX (referred to throughout this manuscript as siTOX) was
purchased from Dharmacon (Lafayette, CO, USA). Dulbecco’s modified eagle
medium (DMEM), minimum essential medium (MEM), fetal bovine serum (FBS),
penicillin/streptomycin, and phosphate buffered saline (PBS) from Gibco,
Invitrogen (UK). Cervical cancer cell line HeLa (CCL-2.2) and human lung
carcinoma A549 (CCL-185) from ATCC (UK). DeadEnd™ Fluorometric TUNEL
System was from Promega (UK). Annexin-V-Fluos staining kit from Roche
Diagnostics GmbH (Germany). Six- to eight-week-old BALB/c mice were obtained
Appendix B: experimental section
124
from Harlan (Oxfordshire, UK), allowed to acclimatise for 1 week and were kept in
groups of 5 for the duration of the experiments and given food and water. All
experiments were conducted with prior approval from the UK Home Office.
Standard dissociated hippocampal cultures were prepared according to
Malgaroli1. Hippocampi were dissected from 0- to 3-day-old animals killed by
decapition in accordance with the regulations of the Italian Animal Welfare Act and
approved but the local Authority Veterinary Service.
B.2 Syntheses
Synthesis of G0 dendron-MWCNT 1. 300 mg of MWCNTs were suspended in 250
mL of DMF. The mixture was first sonicated in sonic bath for 5 minutes, and then
paraformaldehyde (450 mg) and {2-[2-(2-tert-butoxycarbonylamino-ethoxy)-
ethoxy]-ethylamino}-acetic acid (450 mg) were added and the addition was
repeated every 24h for three times, sonicating 15 minutes each time. The reaction
was heated at 115°C for 120h. The product was filtered with a Millipore system (JH
0.45 μm filter) dried with diethyl ether and recovered, sonicated for 30 minutes in
DMF, then filtered and recovered again and finally sonicated for 30 minutes in
MeOH. Following a final filtration, the product, obtained as a fine black powder
(339 mg), was dried at the vacuum pump. TEM charcterization of Boc-protected
MWCNT 1 is reported at pag. 32, AFM characterization is reported at pag. 33 and
the average diameter ranges from 20 to 25 nm. The Kaiser test of this derivative
was 0 mol/g, the loss of weight at 450°C measured by TGA was 3.7% in the batch
reported in Chapter 1 and 1.2% for the batch reported in Chapter 2.
Boc-protected G0 dendron-MWCNT 1 (240 mg) underwent deprotection via
HClg in 200 mL of DMF, stirring at RT overnight. Again, the product was filtered
with a Millipore system (JH 0.45 μm filter), washed with ether and recovered,
sonicated for 30 minutes in DMF, then filtered and sonicated again for 30 minutes
in MeOH. Afterwards, 230 mg of G0 dendron-MWCNT 1 were recovered. The Kaiser
test for the first batch reported in Chapter 2 was 481 mol/g, and 72 mol/g for
Appendix B: experimental section
125
the batch reported in Chapter 3. TEM characterization for MWCNT 1 is reported at
pag. 30.
Synthesis of alkylated G0 dendron-MWCNT 2. 10 mg of MWCNT 1 were added to
5 mg of glycidyl trimethylammonium chloride in 15 mL of MeOH, and the resulting
mixture was kept at 40 °C for 48h. Dendron-MWCNT 2 was filtered with a
Millipore system (JH 0.45 μm filter), washed with diethyl ether and recovered.
Then the suspension was sonicated for 30 minutes in MeOH and washed with
diethyl ether and recovered, then filtered and sonicated again. Following a final
filtration, dendron-MWCNT 2 was obtained as a fine black powder (10 mg). Kaiser
test of MWCNT 2 was 0 mol/g, the loss of weight measured by means of TGA at
450°C was 1.4%. TEM charcterization is reported at pag. 30.
Synthesis of G0.5 dendron-MWCNT 3. 210 mg of MWCNT 1 were mixed with 105
mL of methyl acrylate and 21 mL of N-ethyldiisopropyl amine in 100 mL of MeOH,
and the resulting mixture was kept at 80 °C for 72h. Dendron-MWCNT 3 was
filtered with a Millipore system (JH 0.45 μm filter), and then sonicated with MeOH
and washed with diethyl ether as previously described. 190 mg were recovered.
The Kaiser test of MWCNT 3 was 0 mol/g.
Synthesis of G1 dendron-MWCNT 4. 180 mg of MWCNT 3 were added to 90 mL of
diethylamine and 90 mL of MeOH, the reaction was stirred at 80°C for 72h.
Dendron-MWCNT 4 was filtered with a Millipore system (JH 0.45 μm filter), and
then sonicated with MeOH and washed with diethyl ether as previously described.
After a final filtration, dendron-MWCNT 4 was recovered as a fine black powder
(190 mg). The Kaiser test of MWCNT 4 was 0 mol/g and the relative loss of
weight calculated by means of TGA was 12.5% for the bactch studied in Chapter 1.
The Kaiser test for the bactch studied in Chapter 2 was 111 mol/g.
Synthesis of alkylated (G1) dendron-MWCNT 5. 10 mg of MWCNT 3 were added
to 7.6 mg of glycidyl trimethylammonium chloride in 15 mL of MeOH, and the
resulting mixture was kept at 40 °C for two days. Dendron-MWCNT 5 was filtered
Appendix B: experimental section
126
with a Millipore system (JH 0.45 μm filter), and then sonicated with MeOH and
washed with diethyl ether as previously described. After a final filtration, dendron-
MWCNT 5 was obtained as a fine black powder (10 mg). The Kaiser test of
derivative MWCNT 5 was 0 mol/g, the loss of weight calculated by TGA 3.8%.
TEM characterization is reported at pag. 30.
Synthesis of G1.5 dendron-MWCNT 6. 178 mg of MWCNT 5 were mixed with 60
mL of methyl acrylate and 18 mL of N-ethyldiisopropyl amine in 60 mL of MeOH,
and the resulting mixture was kept at 80 °C for 72h. The product was filtered with
a Millipore system (JH 0.45 μm filter), and then sonicated with MeOH and washed
with diethyl ether as previously described. 173 mg were recovered. Kaiser test for
MWCNT 6 was 0 mol/g. TEM characterization is reported at pag. 32.
Synthesis of (G2) dendron-MWCNT 7. 130 mg of MWCNT 6 were added to 65 mL
of diethylamine and 65 mL of MeOH, the reaction was stirred at 80°C for 72h.
Dendron-MWCNT 7 was filtered with a Millipore system (JH 0.45 μm filter), and
then sonicated with MeOH and washed with diethyl ether as previously described.
After a final filtration, dendron-MWCNT 7 was obtained as a fine black powder
(134 mg). Kaiser test for MWCNT 7 was 1097 mol/g and loss of weight was 18%
for the batch reported in Chapter 1. TEM characterization is reported at pag. 31.
For the batch reported in Chapter 2, Kaiser test was 212 mol/g.
Synthesis of alkylated G2 dendron-MWCNT 8. 70 mg of MWCNT 7 were added to
755 mg of glycidyl trimethylammonium chloride in 70 mL of MeOH, and the
resulting mixture was kept at 40 °C for two days. Dendron-MWCNT 8 was filtered
with a Millipore system (JH 0.45 μm filter), and then sonicated with MeOH and
washed with diethyl ether as previously described. After a final filtration, dendron-
MWCNT 8 was obtained as a fine black powder (72 mg). Kaiser test for MWCNT 8
was 0 mol/g, the loss of weight was 28% in the batch described in Chapter 1 and
5.6% for the second batch reported in Chapter 2. TEM characterization was
reported at pag. 30, 31 and 61. AFM characterization is reported at pag. 33 and the
Appendix B: experimental section
127
average diameter ranges from 35 to 40 nm. E-EFM characterization is reported at
pag. 37.
Synthesis of SWCNT 9. 8 mg of HiPCO purified SWCNTs, used as received from
NRI®, were dispersed in 8 mL of DMF and mixed with sarcosine (12 mg) and
heptaldehyde (12 mg), which were added other three times every 24h and
sonicated for 15 minutes each, for 120h at 120°C. Afterwards it was filtrated
through a Millipore filter (JH 0,45 mm). The resulting material was washed on the
filter with DMF and diethylether, recovered, and re-suspended in DMF upon
sonication for 20 minutes. It was re-filtered, re-washed on the filter with DMF and
diethylether and finally 6 mg were recovered. TGA loss of weight calculated at
500°C was 7%. TEM charcterization is reported at pag. 46, UV-Vis-NIR spectrum is
reported at pag. 47.
Synthesis of SWCNT 10. 5 mg of HiPCO purified SWCNTs, used as received from
NRI®, were dispersed in 0.5 mL of [bmim]BF4) and mixed with sarcosine (7.5 mg)
and aldehyde (7.5 mg), added othe three times every 24h, sonicated for 15 minutes
each, and heated for 120h at 120°C. Afterwards they were filtrated through a
Millipore filter (JH 0,45 mm). The resulting material was washed on the filter with
DMF and diethylether and recovered, then it was re-suspended in DMF upon
sonication for 20 minutes. It was re-filtered, re-washed on the filter with DMF and
diethylether and at this point the collected black material was suspended in
acetone and sonicated for 20 minutes in order to remove all the ILs. Finally, it was
filtered again, washed on the filter with acetone and diethylether. 6 mg of product
were recovered. TGA loss of weight calculated at 500°C was 7%. TEM
charcterization is reported at pag. 48, UV-Vis-NIR spectrum is reported at pag. 46.
Synthesis of SWCNT 11. 5 mg of HiPCO purified SWCNTs, used as received from
NRI®, were dispersed in the IL ([bmim]BF4) and mixed with sarcosine (1 mmol)
and aldehyde (2 mmol). The reactions were carried out in closed glass test tubes
with stirring, under microwaves irradiation (20 W at 120°C). The reaction
Appendix B: experimental section
128
mixtures were cooled down at room temperature and then suspended in DMF.
Afterwards they were filtrated through a Millipore filter (JH 0,45 mm). The
resulting material was washed on the filter with DMF and diethylether and
recovered, then it was re-suspended in DMF upon sonication for 20 minutes. It was
re-filtered, re-washed on the filter with DMF and diethylether and at this point the
collected black material was suspended in acetone and sonicated for 20 minutes in
order to remove all the ILs. Finally, it was filtered again, washed on the filter with
acetone and diethylether. About 5 mg were recovered. Thermal reactions, with and
without ILs, were carried out for comparison purposes, by heating analogue
mixtures in closed vials using an oil bath. TGA loss of weight calculated at 500°C
was 9%. Raman spectrum was reported at pag. 47. TEM charcterization is reported
at pag. 48, UV-Vis-NIR spectrum is reported at pag. 47.
Synthesis of SWCNT 12. 5 mg of SWCNTs were dispersed in 0.5 mL of [omim]BF4,
afterwards 1 mmol of sarcosine and 2 mmol of heptaldehyde were added to the
mixture. The reaction was carried out under MW irradiation at 20 W, for 1 hour
(Tbulk= 120°C), with stirring. After the work-up procedure as described abovefor
compund 11, 5 mg of product were recovered. TGA loss of weight calculated at
500°C was 6%. TEM charcterization is reported at pag. 46, UV-Vis-NIR spectrum is
reported at pag. 46.
Synthesis of SWCNT 13. 5 mg of SWCNTs were dispersed in 0.5 mL of
[hvmim][(CF3SO2)2N], afterwards 1 mmol of sarcosine and 2 mmol of
heptaldehyde were added to the mixture. The reaction was carried out under MW
irradiation at 20 W, for 1 hour (Tbulk= 120°C), with stirring. After the work-up as
described above for compound 11, 6 mg of product were recovered. TGA loss of
weight calculated at 500°C was 0%. TEM charcterization is reported at pag. 46, UV-
Vis-NIR spectrum is reported at pag. 46.
Synthesis of SWCNT 14. 5 mg of SWCNTs where dispersed in 0.5 mL of
[omim]BF4 and 170 L of o-DCB, afterwards 1 mmol of sarcosine and 2 mmol of
Appendix B: experimental section
129
heptaldehyde were added to the mixture. The reaction was carried out under MW
irradiation at 20 W, for 1 hour (Tbulk= 120°C), with stirring. After the work-up as
described above for compound 11, 6 mg of product were recovered. TGA loss of
weight calculated at 500°C was 16%. Raman spectrum was reported at pag. 47.
TEM charcterization is reported at pag. 48, UV-Vis-NIR spectrum is reported at
pag. 47.
Synthesis of SWCNT 15. 5 mg of SWCNTs were dispersed in 0.5 mL of [bmim]BF4,
yielding a gel, afterwards equimolar amounts (0.5 mmol) of sarcosine and
C8F17(CH2)2CHO were added to the mixture. The reactions were carried out under
MW irradiation at 20 W for 1 hour (Tbulk= 120°C), with stirring. After the work-up
as described above for compound 11, 6 mg of product were recovered. TGA loss of
weight calculated at 500°C was 14%. TEM charcterization is reported at pag. 49
and 50, UV-Vis-NIR spectrum is reported at pag. 50.
Synthesis of SWCNT 16. 5 mg of SWCNTs were dispersed in 0.5 mL of [bmim]BF4,
yielding a gel, afterwards equimolar amounts (0.2 mmol) of sarcosine and
C8F17(CH2)2CHO were added to the mixture. The reactions were carried out under
MW irradiation at 20 W for 1 hour (Tbulk= 120°C), with stirring. After the work-up
as described above for compound 11, 6 mg of product were recovered. TGA loss of
weight calculated at 500°C was 16%. TEM charcterization is reported at pag. 49
and 50, UV-Vis-NIR spectrum is reported at pag. 50.
Synthesis of SWCNT 17. 5 mg of SWCNTs were dispersed in 0.5 mL of [bmim]BF4,
yielding a gel, afterwards equimolar amounts (0.15 mmol) of sarcosine and
C8F17(CH2)2CHO were added to the mixture. The reactions were carried out under
MW irradiation at 50 W for 1 hour (Tbulk= 140°C), with stirring. After the work-up
as described above for compound 11, 6 mg of product were recovered. TGA loss of
weight calculated at 500°C was 27%. TEM charcterization is reported at pag. 49
and 50, UV-Vis-NIR spectrum is reported at pag. 50.
Appendix B: experimental section
130
Synthesis of DTPA-G2 dendron-MWCNT 7. MWCNT 7 (5 mg) were dispersed in
dry DMSO/DMF (0.4:2.5 mL) and neutralized with DIEA (diisopropylethylamine) (50
l) to (MWCNT 7). Diethylentriaminepentaacetic (DTPA) dianhydride (28.5 mg) was
added and the mixture was stirred for 48 h at room temperature. The DMSO solution
was partly evaporated and diluted with water. The MWCNTs were recovered by
centrifugation, re-dispersed in water and lyophilized. 5 mg were recovered. The Kaiser
test was 0 mol/g and TEM characterization was reported at pag. 70.
Preparation of [111In]DTPA-MWCNT 7. As a standard procedure, dispersions of
DTPA-MWCNT or DTPA alone (80-400 µL of 250 µg/mL) were diluted with an
equal volume of 0.2 M ammonium acetate buffer pH 5.5, to which 2-20 MBq as
indium chloride (111InCl3) was added. The indium was left to react with the DTPA-
MWCNT and DTPA alone for 30 min at room temperature, after which the reaction
was quenched by the addition of 0.1 M EDTA chelating solution (1/20 the reaction
volume is added). 111InCl3 alone, used as a control, was also subjected to the same
conditions of the labeling reaction. Aliquots of each final product were diluted five
folds in PBS and then 1 µL spotted on silica gel impregnated glass fiber sheets
(PALL Life Sciences, UK). The strips were developed with a mobile phase of 50 mM
EDTA in 0.1 M ammonium acetate and allowed to dry before analysis. This was
then developed and the autoradioactivity quantitatively counted using a Cyclone
phosphor detector (Packard Biosciences, UK). The immobile spot on the TLC strips
indicated the percentage of radiolabeled [111In]DTPA-MWCNT conjugate, while the
free 111In or [111In]DTPA were shown by the mobile spot. The mobile spot on the
TLC strips indicated the percentage of free DTPA present in the DTPA-MWCNT
mixture while precipitated 111In or radiolabeled [111In]DTPA-MWCNT conjugate
remain at the application point.
Synthesis of MWCNT 9. 50 mg of MWCNTs, used as received, were dispersed in 8
mL of DMF and mixed with sarcosine (75 mg) and heptaldehyde (75 mg), added
other three times every 24h, sonicated for 15 minutes each, and then heated for
120h at 120°C. Afterwards it was filtrated through a Millipore filter (JH 0,45 mm).
Appendix B: experimental section
131
The resulting material was washed on the filter with DMF and diethylether and
recovered, then it was re-suspended in DMF upon sonication for 20 minutes. It was
re-filtered, re-washed on the filter with DMF and diethylether and finally 22 mg
were recovered. The metallic residue of MWCNT 9 was 3.9%. TEM characterization
is reported at pag. 78.
B.3 Characterization techniques and sample preparation
Thermogravimetric analysis. TGA Q500 (TA Instruments) was used to record
TGA analysis under N2 or under air, by equilibrating at 100°C, and following a
ramp of 10 °C/min up to 1000°C. About 0.5 mg per each analysis were required.
Transmission Electron Microscopy (Trieste). TEM analyses were performed on
a TEM Philips EM208, using an accelerating voltage of 100 kV. About 1 mg of each
compound was solubilized or dispersed in 1 mL of solvent. Then, one drop of this
solution was deposited on a TEM grid (200 mesh, Nichel, carbon only; carbon
lacey).
Transmission Electron Microscopy (London). Dendron-MWCNT 8 (15 µL of 1.0
mg/mL in 10% dextrose) was complexed with an equal volume of siRNA at 1:16
siRNA:dendron-MWCNT mass ratio. The complex was left to equilibrate for 30 min.
A drop of the suspension was placed on a grid with a support film of
Formvar/carbon, excess material was blotted off with a filter paper and the
complexes were examined under a FEI CM120 BioTwin Transmission Electron
Microscope (Einhoven, Netherlands) using a Lab6 emitter. Images were captured
using an AMT Digital Camera.
Atomic force microscopy. AFM samples were prepared by the deposition of one
drop of a highly diluted nanotubes solution on mica substrate and spin coating in a
row. Pristine MWCNTs and MWCNTs after cycloaddition were dispersed in DMF,
MWCNT 8 was dispersed in MeOH. The samples were examined with a XE-100
microscope (Park Scientific Instruments Advanced Corp.–PSIA) at room
Appendix B: experimental section
132
temperature. In all AFM measurements, the images were taken in non-contact (NC)
mode with scan speed of 0.3 Hz using standard silicon rectangular cantilevers
(Park, 950M-ACTA) with nominal spring constants of 40 N/m and radius of
curvature < 10 nm. In all NC-AFM measurements, the tip scanned the sample with
the drive frequency of 330 kHz to acquire the topographic image of various sizes
(range from 1.2 µm to 5 µm). The distributions of average diameters were
obtained by estimating diameter of single MWCNTs (with and without
functionalizations) from NC-AFM topographic images and their corresponding
average line profiles. A number of 36, 24 and 37 nanotubes were observed for
evaluating the distribution of average diameters of pristine MWCNTs, conjugates
boc-protected 1 and 8, respectively.
Electrostatic Force Microscopy. The EFM analysis was performed using a PSIA
XE-100 AFM endowed with an external lock-in amplifier (SR830, Stanford
Research Systems) in order to carry out single pass EFM characterization of
surfaces. A solution of pristine MWCNTs in DMF (1 mg/mL) or of functionalized
CNT in MeOH (1 mg/mL) were put for 3 times (50 L each time, total volume 150
L) on a silicon wafer (2 cm x 1 cm) by spin coating (6000 rpm for 15 sec each).
The spin coater was a Laurell Technologies Corporation 10:04 AM Model WS-
400B_6NPP/LITE model.
Bucky gel preparation (Chieti). The proper amount of purified and unpurified
HiPCO SWCNTs was poured in an agate mortar and grinded for 30 min with 1 mL
(or 1g in the case of solid IL) of the ionic liquid. The obtained suspension was
poured in a vessel and sonicated with a ultrasonic bath sonicator (Bandelin
Sonorex, 35 KHz) for 1 h. The sonication increases the apparent viscosity of
originally liquid samples. [Hvim][(CF3SO2)2N] is solid at room temperature,
therefore, the grey powder obtained by 30 min grinding in the mortar was poured
in a vessel and heated at 50°C before sonication. The obtained samples were either
used for synthetic purposes or poured in the rheometer for the rheological
measurements. No centrifugation step was performed in order to remove excess
Appendix B: experimental section
133
IL. The samples appear to be quite stable and do not tend to display separation
into a black gel phase and a transparent liquid phase within a few weeks.
Rheology (Chieti). Static and dynamic measurements were performed on a
Thermo Scientific modular rheometer Haake M.A.R.S. II equipped with Electron
Corporation thermo-controller system Haake Phoenix (data evaluation: RheoWin
software 3.61) using the cone plate geometry (cone diameter 6 mm, angle 0.5°). All
samples, subjected to vacuum pumping before measurements, were directly
loaded onto the plate of the rheometer and were allowed to equilibrate for 6
minutes. As a matter of fact we did not prestress the samples or allow a longer
equilibration time because prestressing the samples or allow equilibration would
have required to leave the samples undisturbed between the plates for ca. 24 h and
this delay would have favoured the undesired hydration of the ionic liquids. The
shear stress sweep tests showed that 1.0 Pa satisfies the linear viscoelasticity for
all samples. The dynamic storage and loss moduli were examined in the linear
viscoelastic regime at 25°C unless otherwise stated.
Microwave. Continuous microwave irradiation was carried out in a CEM-Discover
monomode microwave apparatus, with simultaneous monitoring of irradiation
power, pressure and temperature. Compressed air was applied to improve the
temperature control of the reaction mixtures.
Raman spectroscopy. Micro-Raman spectra were recorded with a Via Renishaw
spectrometer equipped with a He-Ne (632.8 nm) laser. The analyses were
conducted on solid samples.
UV-Vis-NIR spectroscopy. UV-Vis-NIR spectra were recorded in 1-cm quartz
cuvettes on a Varian Cary 5000 spectrophotometer. The spectra were recorded in
DMF solution
Electrophoretic motility shift assay. 0.5 µg of siRNA in 30µL of water complexed
to dendron-MWCNT 8 at different mass ratios, or 0.5 µg of free siRNA as a control,
Appendix B: experimental section
134
were added to 1% agarose gel containing ethidium bromide. The gel was run for
45 min at 70 V and then photographed under UV light using GeneGenius system,
PerkinElmer Life and Analytical Sciences (USA).
Evaluation of fluorescent siRNA complex uptake in HeLa Cells. ATTO 655-
labeled siRNA (ex= 655 nm, em= 690 nm) was used to prepare fluorescent siRNA
complexes. siRNA was complexed with the different generations of dendron-
MWCNTs at a 1:16 mass ratio. HeLa cells were grown to confluency in glass
coverslips in 24-well tissue culture dishes at a density of 50000 cells per well in
MEM supplemented with 10% FBS and 1% penicillin/streptomycin. After 24h, the
cells were trasfected with non-coding siRNA conjugated to ATTO 655. Briefly, 50
L of the preformed siRNA complex was diluted 10 times with complete medium,
and 500 L of the complex-containing medium was added to each well yielding a
final siRNA concentration of 80 nM (1g/mL). The cells were incubated with the
siRNA:MWCNT complexes for 4h in serum-free medium, after which serum was
added to achieve a final serum concentration of 10%. After 24h, the cells were
washed with PBS solution, fixed with 4% paraformaldehyde in PBS for 15 minutes
at RT, and then rinsed with PBS. For nuclear staining, cells were permeabilized
with 0.1% Triton X-100 in PBS for 10 minutes at RT, treated with RNAase (100
g/mL) for 20 minutes at 37°C, incubated with 1 mg/mL propidium iodide
(Molecular Probes) in PBS for 5 minutes, and then rinsed three times with PBS.
Coverslips were mounted with aqueous poly(vinyl alcohol) Citifluor reagent mixed
with AF100 antifade reagent (10:1) (Citifluor) prior to use. Slides were examined
under the CLSM using a 63X oil immersion lens.
Assessment of cell viability. A549 cells (50,000 cells/well) maintained in DMEM
media supplemented with 10 % fetal bovine serum (FBS), 50 U/mL penicillin, 50
μg/mL streptomycin, 1% L-glutamine and 1% non-essential amino acids at 37 °C
in 5% CO2, were sub-cultured into 24-well plates for 24h before incubation with
the f-CNT. Cells were incubated with dendron-MWCNT 8 in serum-free media for
Appendix B: experimental section
135
4h after which serum-containing media was added to make the final concentration
of serum to 10%. Cell death was assessed after 24h of incubation (immediate) or
allowed to recover (delayed) for additional 48h. Final concentrations reached were
10 µg/mL. The live-death cell assay was performed with a PI/annexin-V-FITC
staining kit according to the instructions of the manufacturer. A CyAn ADP flow
cytometer (DakoCytomation) was used to analyze 20000 cells per sample. All
conditions were tested in triplicate.
Transfection of HeLa cells with siRNA. HeLa cells were maintained in MEM
media, supplemented with 10% fetal bovine serum (FBS), 50 U/mL penicillin, 50
μg/mL streptomycin, 1% L-glutamine and 1% non-essential amino acids at 37°C in
5 % CO2. Cells (5000 cells/well) were subcultured into 16-well chamber slides.
24hs later, cells were transfected with siTOX, an apoptosis inducing siRNA. Briefly,
30 µL of the pre-formed dendron-MWCNT:siRNA complex was diluted 5 times with
serum free media and 150 µL of the complex containing media were added to each
well yielding a final siRNA concentration of 80 nM. Four hours later, 150 µL of
fresh media containing 20% FBS was added to each well. After 48 h incubation at
37°C and 5 % CO2, the DeadEnd™ Fluorometric TUNEL System was used to label
nicked DNA through incorporation of fluorescein-12-dUTP. Samples were
incubated with recombinant terminal deoxynucleotidyl transferase (rTdT)
indicated by manufacturer instructions and fluorescein labelling was visualized
using confocal laser scanning microscopy (CLSM) (Zeiss LSM 510 Meta; Carl Zeiss,
Oberkochen, Germany) using 30 mW 488 nm Argon laser excitation source, LP 505
nm output filter, and a Plan-Neofluar 10X lens. Propidium iodide was used to
counterstain nuclei.
Whole body imaging of [111In]DTPA-MWCNT using Nano-SPECT/CT. Balb/C
mice were anaesthetised by isofluorane inhalation. Each animal was injected via
tail vein injection with 250 µL containing 50 µg of [111In]DTPA-MWCNT containing
approximately 5-6MBq. [111In]EDTA was injected for comparison. Immediately
after injection (t=0.5hrs) and at 3.5hs and 24hs, mice were imaged using the Nano-
Appendix B: experimental section
136
SPECT/CT scanner (Bioscan, USA). SPECT images were obtained in 16 projections
over 40-60 min using a 4-head scanner with 1.4 mm pinhole collimators. CT scans
were taken at the end of each SPECT acquisition and all images were reconstructed
with MEDISO software (Medical Imaging Systems). Fusion of SPECT and CT images
was carried out using the PMOD software.
Tissue biodistribution of [111In]DTPA-MWCNT. Animals (n=3-4) were injected
with 250µL containing 50 µg of [111In]DTPA-MWCNT or [111In]EDTA in 5%
dextrose via tail vein injection. Twenty four hours after injection, mice were killed,
and blood was collected. Heart, lungs, liver, spleen, kidneys, muscle and bone were
sampled, each sample being weighed and counted on a Gamma Counter (Perkin
Elmer, USA), together with a dilution of the injected dose with dead-time limit
below 60%. The percentage injected dose per gram tissue or the percentage
injected dose per organ was calculated for each tissue.
Preparation of CNT coverslips from a water solution. 5 mL of a water solution
(0.04 mg/mL) of MWCNTs 9 were evaporated over 3 glass substrates, those were
deposited in a crystallizer (crystallizer surface area = 1809 mm2). The solution was
slowly dried out. The substrates were then placed in oven at 350 °C under nitrogen
atmosphere in order to remove the organic functionalization.
Preparation of CNT coverslips from a DMF solution. The coverslips were
heated at 80°C and 2 mL (200 L x 10 times) of a 0.01 mg/mL solution of
MWCNTs 9 were deposited onto their surface. Then, we treated the substrates for
20 minutes at 350°C under N2 atmosphere. SEM characterization of these
substrates is reported at pag. 80 and 81.
Preparation of neuronal cultures1. Hippocampi were isolated from newborn
rats, digested and recovered, cells were plated on bare glass coverslips and on
CNTs treated glass coverslips. Coverslips were placed in Petri dishes and cultured
in serum containing medium in a 5% CO2-humified incubator for 7-9 days.
Appendix B: experimental section
137
Scanning Electron Microscopy. SEM measurements were carried out with a Zeiss
Supra™ microscope. Substrates with cells were metalized with about 10 nm of Ni
(6 seconds, 300 W, throttle valve 25%, Argon 25 sccm).
Electrophysiological recordings. Cultures were superfused with a solution
containing (mM): NaCl 150, KCl 4, MgCl2 1, CaCl2 2, HEPES 10, glucose 10.
Recordings were obtained from neurons with patch pipettes (4-7 M) using an
Axoclamp 2 unit (Axon Instrument) in current clamp mode. Pipettes were filled
with a solution containing (mM): K gluconate 120, KCl 20, HEPES 10, EGTA 10,
MgCl2 2, Na2ATP 2 (pH 7.35 adding KOH). In voltage clamp an EPC-7 Amplifier
(List Germany) was used, and all recordings were taken at the same holding
potential of -58 mV. Membrane potentials were corrected for liquid junction
potentials.
References:
(1) Malgaroli, A.; Tsien, R. Glutamate-induced long-term potentiation of the
frequency of miniature synaptic currents in cultured hippocampal neurons.
Nature 1992, 357, 134-139.
top related